Lithos 97 (2007) 336 – 364
www.elsevier.com/locate/lithos
Interrelations between coeval mafic and A-type silicic magmas from
composite dykes in a bimodal suite of southern Israel, northernmost
Arabian–Nubian Shield: Geochemical and isotope constraints
Y. Katzir a , B.A. Litvinovsky a,⁎, B.M. Jahn b , M. Eyal a , A.N. Zanvilevich a ,
J.W. Valley c , Ye. Vapnik a , Y. Beeri a , M.J. Spicuzza c
a
Department of Geological and Environmental Sciences, Ben Gurion University of the Negev, Beer Sheva 84105, Israel
b
Institute of Earth Sciences, Academica Sinica, Taipei 11529, Taiwan
c
Depatment of Geology and Geophysics, University of Wisconsin, Madison WI 53706, USA
Received 30 September 2006; accepted 3 January 2007
Available online 13 January 2007
Abstract
Late Neoproterozoic bimodal dyke suites are abundant in the Arabian–Nubian Shield. In southern Israel this suite includes
dominant alkaline quartz porphyry dykes, rare mafic dykes, and numerous composite dykes with felsic interiors and mafic margins.
The quartz porphyry chemically corresponds to A-type granite. Composite dykes with either abrupt or gradational contacts between
the felsic and mafic rocks bear field, petrographic and chemical evidence for coexistence and mixing of basaltic and rhyolitic
magmas. Mixing and formation of hybrid intermediate magmas commenced at depth and continued during emplacement of the
dykes. Oxygen isotope ratios of alkali feldspar in quartz porphyry (13 to 15‰) and of plagioclase in trachydolerite (10–11‰) are
much higher than their initial magmatic ratios predicted by equilibrium with unaltered quartz (8 to 9‰) and clinopyroxene (5.8‰).
The elevation of δ18O in alkali feldspar and plagioclase, and extensive turbidization and sericitization call for post-magmatic lowtemperature (≤ 100 °C) water–rock interaction. Hydrous alteration of alkali feldspar, the major carrier of Rb and Sr in the quartz–
porphyry, also accounts for the highly variable and unusually high I(Sr) of 0.71253 to 0.73648.
The initial 143Nd/144Nd ratios, expressed by εNd(T) values, are probably unaltered and show small variation in mafic and felsic
rocks within a narrow range from + 1.4 to + 3.3. The Nd isotope signature suggests either a common mantle source for the mafic
and silicic magmas or a juvenile crustal source for the felsic rocks (metamorphic rocks from the Elat area). However, oxygen
isotope ratios of zircon in quartz porphyry [δ18O(Zrn) = 6.5 to 7.2‰] reveal significant crustal contribution to the rhyolite magma,
suggesting that mafic and A-type silicic magmas are not co-genetic, although coeval. Comparison of 18O/16O ratios in zircon
allows to distinguish two groups of A-type granites in the region: those with mantle-derived source, δ18O(Zrn) ranging from 5.5 to
5.8‰ (Timna and Katharina granitoids) and those with major contribution of the modified juvenile crustal component, δ18O(Zrn)
varying from 6.5 to 7.2‰ (Elat quartz porphyry dykes and the Yehoshafat alkaline granite). This suggests that A-type silicic
magmas in the northern ANS originated by alternative processes almost coevally.
© 2007 Elsevier B.V. All rights reserved.
Keywords: Composite dyke; A-type granite; Oxygen isotopes; Sr isotopes; Nd isotopes; Arabian–Nubian shield
⁎ Corresponding author.
E-mail address: [email protected] (B.A. Litvinovsky).
0024-4937/$ - see front matter © 2007 Elsevier B.V. All rights reserved.
doi:10.1016/j.lithos.2007.01.004
Y. Katzir et al. / Lithos 97 (2007) 336–364
1. Introduction
Composite dykes that are made up of felsic and mafic
rocks are abundant worldwide (e.g., Vogel and Wilband,
1978; Marshall and Sparks, 1984; Furman and Spera,
1985; Stern and Voegeli, 1987; Carrigan and Eichelberg,
1990; Poli and Tommazini, 1991; Koyaguchi and
Takada, 1994; Litvinovsky et al., 1995a; Titov et al.,
2000; Jarrar et al., 2004). Composite dykes, like
microgranular mafic enclaves (MME) and synplutonic
basic dykes in felsic igneous rocks, provide unequivocal
evidence for coexistent magmas of contrasting composition. Comparative study of mafic and felsic rocks in
composite dykes can shed light on possible genetic links
between mafic magmas and contemporaneous silicic
melts. In post-orogenic settings the felsic rocks of
composite dykes exhibit geochemical traits of A-type
granite, so the felsic rocks may serve as one of key
objects for study of the origin of A-type granites (Stern
and Voegeli, 1987; Jarrar et al., 2004).
Two different types of composite dykes are distinguished (Snyder et al., 1997; Wiebe and Ulrich, 1997):
(Type 1) those with felsic margins and mafic interiors;
the latter are commonly in the form of pillow-like and
rounded enclaves or as a net-veined essentially mafic
dykes; (Type 2) dykes with felsic interiors and mafic
margins; a main distinctive feature of these dykes is
clear zoning with smooth planar contacts, both abrupt
and gradational, between felsic interior and mafic
marginal zones.
Type 1 dykes are less appropriate for reconstruction
of the initial magma compositions due to the high
extent of interaction between the commingling magmas. This particularly affects the more mobile elements
such as K, Rb, Na, Ba, Ca, Sr (Zorpi et al., 1991; Poli
and Tommazini, 1991; Litvinovsky et al., 1995b). It is
also known that exchange of radiogenic isotopes between
two co-eval magmas occurs much faster than element
diffusion (Snyder and Tait, 1998; and references therein).
This causes ambiguity in the study of consanguinity of
coeval mafic and silicic melts and sources of mafic
magmas.
Type 2 composite dykes provide more reliable
information on the chemical composition of coexisting
magmas. The dykes are thought to form in shallow level
plutonic or subvolcanic systems when basaltic melts
ruptured a granitic magma reservoir and injected country
rocks forming dykes outside the reservoir; silicic magma
flowed outward in the center of basaltic dyke prior to its
complete solidification (Gibson and Walker, 1963;
Kanaris-Sotiriou and Gill, 1985; Wiebe and Ulrich,
1997). Depending on the time interval between the two
337
successive injections, contacts between felsic interiors
and mafic margins vary from clearly gradational to
abrupt. Abrupt contacts, common in many dykes, suggest that interaction of two magmas was inhibited due to
the high extent of crystallization of the earlier mafic
magma. In composite dykes with gradational contacts
the initial composition of mafic rocks could be preserved at least in the outermost parts where the injected
mafic melt solidified rapidly adjacent to cold country
rocks.
Composite dykes of type 2 are abundant in the
northernmost part of the Arabian–Nubian shield
(ANS), in particular in the Eastern Desert and Sinai
(Egypt), in southwestern Jordan and in southern Israel
(Stern and Voegeli, 1987, 1995; Beyth et al., 1994;
Iacumin et al., 1998; Mushkin et al., 1999, 2003; Jarrar
et al., 2004). These dykes are incorporated in bimodal
basalt–rhyolite dyke suites forming dyke swarms up to
tens of kilometers long. The felsic rocks are chemically
referred to as typical A-type granite (Stern et al., 1988;
Kessel et al., 1998; Mushkin et al., 2003). Dykes are
shown to be genetically linked to bimodal volcanic
suites of similar composition (Stern et al., 1988; Jarrar
et al., 1992).
Here we present new data on the structure, and the
geochemical and isotope characteristics of the trachydolerite–quartz porphyry bimodal dyke suite in
southern Israel including composite dykes both with
abrupt and gradational contacts between felsic and
mafic zones. Particular emphasis has been placed on
the magma mixing process and its effect on the original
chemical composition of the rocks. Comparison of
oxygen and Sr isotope data suggests that the Rb–Sr
isotope system had been significantly disturbed by
low-temperature water–rock interaction and, consequently, the measured 87 Sr/86 Sr ratios do not characterize the magmatic process. Despite having similar
εNd(T) the felsic and mafic rocks are not cogenetic:
δ18O values of zircon indicate significant contribution
of upper crustal sources to the A-type rhyolite magma.
2. Geological setting
The study was performed in the Elat area, the largest
exposure of Precambrian basement in southern Israel.
Along with the adjacent Amram and Timna exposures, as
well as a number of outcrops in Jordan, the Elat area
represents the northernmost edge of the ANS (Fig. 1, Inset).
The shield formed during the Neoproterozoic (ca. 900–
550 Ma) as part of the East African Orogen by accretion of
juvenile arc terranes. The accretion occurred at ∼700 Ma
(Stern, 1994 and references therein) and was followed by
338
Y. Katzir et al. / Lithos 97 (2007) 336–364
continental collision that resulted in magmatic thickening
by intrusion of large calc-alkaline granitic batholiths
(Stoeser and Camp, 1985). The batholitic stage (stage 3
of Bentor, 1985) lasted from ∼650 Ma until 620–610 Ma.
The stage ended with regional uplifting and erosion at
∼ 610–600 Ma and formation of molasse deposits
(Garfunkel, 2000; Jarrar et al., 2003). Late Neoproterozoic
crustal evolution (ca. 610–540 Ma) in the ANS occurred in
an extensional tectonic environment (Stern and Hedge,
1985; Bentor and Eyal, 1987; Stern et al., 1988; Baer et al.,
1994; Garfunkel, 2000; Genna et al., 2002) and was
accompanied by abundant igneous activity: emplacement
of alkaline granitoids, eruption of bimodal volcanics, and
intrusion of dyke swarms, also mostly bimodal (Stern et al.,
1984; Stern and Gottfried, 1986; Beyth et al., 1994; Jarrar
et al., 1992, 2004; Stein et al., 1997; Jarrar, 2001; Moghazi,
2002).
In southern Israel basement outcrops comprise metamorphic and magmatic rocks that represent all of the main
stages of the ANS evolution (Bentor, 1985; Garfunkel
et al., 2000). The Late Neoproterozoic stage included
deposition of molasse (Elat conglomerate), formatiom of
bimodal alkaline plutonic complexes (Timna and Amram
complexes) and andesite–rhyolite volcanism (for geological map see Supplementary data, Fig. 1S).
Four successive dyke episodes were widely exhibited
in southern Israel during the second half of the
Neoproterozoic, from ∼ 615 Ma to 530 Ma (Katzir
et al., 2006): (1) Microdiorite and lamprophyre dyke
suite; (2) Dacite porphyry with subordinate andesite and
rhyolite porphyry; (3) Bimodal dyke suite, rhyolites
with small amount of trachydolerite and intermediate
rocks, mostly in composite dykes; (4) Dolerite,
commonly high-Ti. Dyke episodes 1, 2 and 3 were
separated in time by plutonic, volcanic and sedimentation events. The first two suites are timed to the very end
of the batholitic stage; they are overlain by Elat
conglomerate formed at ∼ 600 Ma (Garfunkel, 2000).
Dykes of the bimodal suite and the youngest dolerite
dykes intrude all Neoproterozoic igneous and sedimentary rocks (Kessel et al., 1998; Beyth and Heimann,
1999; Katzir et al., 2006).
The bimodal dyke suite is the most abundant among
the Neoproterozoic dykes in southern Israel (Fig. 1). The
schematic map of the dykes was compiled based on
aerial photograph interpretation; however, field observa-
tions showed that in a number of localities in central and
northern Elat area the number of dykes is an order of
magnitude higher. In the vicinity of Mt. Shelomo, within
an area of about 1 by 2 km they constitute 50 to 80% of
the total rock volume (Gutkin and Eyal, 1998). The
general trend of the dykes is predominantly NS and
changes gradually to N50°E in the northern part of the
area. In the upper reaches of Wadi Shelomo two trends of
dykes are distinguished: N40°W and N30°E. The dykes
are few hundred meters to 3 km long, commonly vertical
and steeply dipping. Their thickness ranges from 1 to
30 m. In larger dykes, en echelon arrayed segments,
commonly 0.5–1 km long, are observed. Felsic rocks,
mostly quartz porphyry, are absolutely dominant in the
suite. The proportion of mafic rocks in the bimodal suite
in southern Israel is approximately 5 vol.%. Trachydolerite and trachyandesite dykes are rare; instead, mafic
rocks constitute margins in composite dykes that are part
of the bimodal dyke suite. About a dozen of such dykes
were found in the Elat area (Fig. 1), and several dykes
were mapped immediately to the north, in the Amram
and Timna exposures (Beyth et al., 1994; Mushkin et al.,
1999). Along-strike tracing of some larger quartz
porphyry dykes often reveals sections with mafic
margins. For instance, in dyke #8 that extends over
2.5 km several sections with mafic margins of about 1 m
wide and 100–600 m long were observed (Fig. 1). This
suggests that the real number of composite dykes may be
larger than the dozen mapped ones.
The composite dykes are mostly subvertical, up to
2.5 km long and commonly 10 to 20 m wide. The felsic
interiors consist of homogeneous quartz porphyry, but in
some dykes they are contaminated by mafic material.
The dyke margins are made up of trachydolerite,
trachyandesite, and trachydacite. In places the composition of the mafic rocks changes along strike from
andesite to dacite. The proportion of mafic rocks in a
composite dyke is generally ≤20%. Contacts between
the felsic interiors and mafic margins are commonly
abrupt, without obvious signs of magma–magma
interaction, but also without evidence of chilling
(Fig. 2A). In few dykes the contact is highly gradational
with heterogeneous mixing between the mafic and
silicic magmas (Fig. 2B); in some dykes abrupt contact
changes along strike to gradational, then back to abrupt
(dyke #1 in Fig. 1).
Fig. 1. Distribution of quartz porphyry, rhyolite porphyry and composite dykes from the bimodal suite in the Elat area (based on an aerial photographs
interpretation performed by V.Voznesensky, with authors' refinements). Area with abundant dykes, after Gutkin and Eyal (1998). Cretaceous to recent
sedimentary sequences (undivided) are shown in white. Inset: 1 and 2 are Late Neoproterozoic bimodal volcanic and dyke suites in southwestern
Jordan (Jarrar et al., 2004) and Northeastern Desert, Egypt (Stern et al., 1988).
Y. Katzir et al. / Lithos 97 (2007) 336–364
339
340
Y. Katzir et al. / Lithos 97 (2007) 336–364
Fig. 2. Composite dykes of the second type in the Elat area. A, sub vertical 20-m wide composite dyke with abrupt contact between felsic interior and
mafic margins. The dyke is exposed in Wadi Shehoret, 5 km to the north from the area shown in Fig. 1 (for locality, see Supplementary data). The left
contact, highlighted by black line is subjected to tectonic fracturing. B, sketch of the structure of a composite dyke with gradational contacts between
felsic interior and mafic margins (Fig. 1, dyke #2). The width of hybrid zones (30–50 cm) is off scale. The inset shows a photograph of typical
microgranular mafic enclaves (MME) in quartz porphyry.
Evidence of commingling and mixing of basaltic and
rhyolitic magmas is pictorially manifested in the
composite Netafim dyke #2 located in the upper reaches
of Wadi Netafim (Fig. 2B). Its thickness ranges from
about 7 to 12 m including trachydolerite margins of 0.4–
0.9 m wide. The contact between the quartz porphyry
interior and trachydolerite margins is gradational, with
two symmetrical transitional zones of about 30–50 cm
in width. The transitional zones, in turn, are divided into
two subzones. In the section adjacent to the quartz
porphyry, the color of the rock progressively changes
from brick red to reddish gray and has trachydacitic
composition. In the section adjacent to the mafic margin
the hybrid rock has dark reddish green color and
trachyandesitic composition. A blurred transition zone
between the hybrid trachydacite and trachyandesite can
be easily traced for a distance of several tens of meters
throughout the outcrop. Quartz porphyry of the interior
of the Netafim dyke contains numerous, but unevenly
distributed fine-grained, oval, rounded and irregular
MME (Inset in Fig. 2B). Enclaves range from few
millimeters to 20–25 cm across and have crenulate,
sinuous, lobate and convex outlines. The curved shape
of some enclaves points to plastic and ductile deformation. Also small mafic schlieren are fairly common
throughout the quartz porphyry. This shows that the
central zone of the Netafim dyke shares some traits of
Type 1 composite dykes.
Y. Katzir et al. / Lithos 97 (2007) 336–364
The stratigraphic position and age of the bimodal
dyke suite is constrained by field relations: dykes
intrude all Late Neoproterozoic plutonic, volcanic and
sedimentary rock units, including a volcanic suite made
up of alkaline rhyolite and andesite (Mushkin et al.,
2003). In turn, they are intersected by dolerite dykes
dated to 531 ± 5 Ma in the Timna area (40Ar/39Ar; Beyth
and Heimann, 1999), and to 545 ± 13 Ma in southwestern Jordan (K–Ar; Jarrar, 2001). The dated dolerite dyke
from Timna is overlain by Early Cambrian sandstone
(Beyth and Heimann, 1999).
Recent U–Pb dating of zircon grains of quartz
porphyry and trachydolerite from composite dyke #2
(Fig. 2B) yielded ages of 592.7 ± 6 Ma and 593.7 ±
8.4 Ma, respectively, which are consistent with the
magmas being coeval (Y. Beeri; in situ ion probe
analysis in the NORDSIM Facility, Swedish Museum of
National History, Stockholm). The analytical procedures
and results will be described in detail in a special paper.
The U–Pb zircon ages are similar to a Rb–Sr whole rock
age of Late Neoproterozoic bimodal swarm and
composite dykes in the Northeastern Desert and Sinai
Peninsula, Egypt (Stern and Hedge, 1985; Stern and
Voegeli, 1987, 1995).
The 593 Ma U–Pb age of the bimodal dyke suite is in
conflict with an age estimate for volcanic rhyolites in
southern Israel, 550 Ma, which is based on a Rb–Sr
isochron (Bielski, 1982). Our study of rhyolites from
the localities sampled by Bielski (Moon Valley,
northern Ramat–Yotam caldera, Amram area) showed
that ultra-K rhyolites with K2O content ranging from
8 to 10 wt.% are widespread in these areas. The highly
unusual composition of the rocks was attributed to
extensive metasomatic alteration that resulted in
dramatic increase of potassium on the expense of
sodium (Agron and Bentor, 1981). Below, in Section
7.4, significant disturbance of the Rb–Sr system, even
in rhyolite dykes with minimal evidence of metasomatic
alteration, is demonstrated. Therefore Rb–Sr dating of
volcanic rhyolite in southern Israel is highly debatable.
Since the dykes from the bimodal suites are considered
as feeders for volcanics (Stern et al., 1988; Jarrar et al.,
1992; Mushkin et al., 1999), the latter cannot be much
older than the dyke rocks.
3. Analytical methods
Whole-rock chemical analyses were performed using
a combination of wet chemical methods, atomic
absorption spectroscopy and titration (major elements),
and X-ray fluorescence (Rb, Sr, Ba, Y, Zr and Nb) at the
Geological Institute, Siberian Division of the Russian
341
Academy of Sciences, Ulan-Ude. Rare earth and some
selected trace elements (Hf, Ta, Th, U, Ga, V, Cu, Pb,
Zn, Sc and Cs) were analyzed by the ICP-MS method at
the National Taiwan University, Taipei and at the
Institute of Mineralogy and Geochemistry of Rare
Elements, Moscow. Analyses are considered accurate
to within 2–5% for major elements, and better than 10–
15% for trace elements. The accuracy for all the REE
(except Lu) is 1–5%; for Lu, it is 9–10%.
Electron microprobe mineral analyses were carried
out using a modernized four-channel MAR-3 electron
probe microanalyser at the Geological Institute, Ulan–
Ude. Analyses were obtained with a beam of 2–3 μm.
Operating conditions were 20 kV, 40 μA beam current,
and a counting time of 10 s. The detection limits are
0.05–0.09 wt.% for Na2O, MgO, Al2O3, and SiO2;
0.01–0.05 wt.% for K2O, CaO, TiO2, MnO and FeO.
For oxygen isotope analysis high purity mineral
separates were hand picked from crushed and sieved
rock samples. Zircon (2.5–3.5 mg per analysis) was
separated from the 63–200 μm size fraction using
standard methods including Wilfley shaking table,
Frantz magnetic separator and heavy liquids. Zircon
concentrates were treated with a series of cold acids to
remove potential contaminants. Purity of zircon separates was ensured by hand picking. Oxygen isotope
analysis of minerals and whole-rock samples was
performed using the laser fluorination technique at the
University of Wisconsin — Madison (Valley et al.,
1995). BrF5 was used as the reagent. Oxygen was
purified cryogenically and with an inline Hg diffusion
pump, converted to CO2 using a hot graphite rod, and
analyzed on a Finnigan MAT 251 mass spectrometer.
Quartz separates (1 to 1.5 mg per analysis) were
analyzed using the rapid heating, defocused beam
technique (Spicuzza et al., 1998a). Plagioclase and
alkali feldspar separates and whole rock powders, which
might react appreciably with BrF5 at room temperature,
were analyzed (2 to 2.5 mg per analysis) using an ‘airlock’ sample chamber, which allows fluorination of
each sample individually (Spicuzza et al., 1998b). The
long-term reproducibilities for conventionally lased
samples and for samples lased in the ‘air-lock’ chamber
in the UW isotope lab are 0.1 and 0.2‰, respectively. On
each day of analysis at least four aliquots of UW Gore
Mountain Garnet standard (UWG-2) were analyzed. The
overall average for 29 analyses on six days of conventionally lased UWG-2 in this study is 5.59 ± 0.13‰.
UWG-2 analyzed using the ‘air lock’ chamber averaged at
5.57± 0.18‰ (n = 36). Data for each day of analysis were
adjusted by an average of 0.21‰ (0.23‰ for ‘air-lock’
analysis days), determined by the difference between each
342
Y. Katzir et al. / Lithos 97 (2007) 336–364
day's UWG-2 value and 5.8‰, the accepted δ18O
(SMOW) value of UWG-2 (Valley et al., 1995).
Rb–Sr and Sm–Nd isotope analyses. Sr and Nd
isotopic ratios were measured using a Finnigan MAT262 thermal ionization mass spectrometer at the Institute
of Earth Sciences (IES), Academia Sinica, Taipei.
Concentrations of Rb, Sr, Sm and Nd were determined
by ICP-MS using an Agilent 7500 s spectrometer at the
National Taiwan University. For concentration determination, fused glass beads were powdered and dissolved
in HF/HNO3 (1:1) mixture in capped Savilex beakers
for N 2 h at ca. 100 °C, followed by evaporation to
dryness, then refluxing in 7N HNO3 for N12 h at ca.
100 °C, and finally, diluting the sample solution by 2%
HNO3. An internal standard solution of 5 ppb Rh and Bi
was added and the spiked solution was further diluted
Table 1
Chemical composition of representative rocks from simple dykes of the bimodal suite (wt.%, ppm)
Sample
no.
A144-2 A149
A157
A150-1 A128
A129 A239 D59
A52
A90-4 D10
D16
D44
A147-2 A156-1 A200 A255
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
SiO2
TiO2
Al2O3
Fe2O3
FeO
MnO
MgO
CaO
Na2O
K2O
P205
LOI
Total
Rb
Sr
Ba
Nb
Ta
Th
U
Zr
Hf
Y
Ga
La
Ce
Pr
Nd
Sm
Eu
Gd
Tb
Dy
Ho
Er
Tm
Yb
Lu
Eu/Eu⁎
Lan/Ybn
(Na + K) /
Al
77.2
0.1
11.7
1.97
0.11
0.06
0.06
0.14
3.41
4.92
–
0.39
100.06
150
27
190
33
7.41
17
5.61
590
13.7
68
21
39
89
10.6
41
9.8
0.18
9.4
1.63
10.3
2.31
6.32
0.95
5.92
0.9
0.06
4.8
0.93
76.7
0.13
11.3
2.05
0.07
0.02
0.14
0.32
3.57
4.86
0.05
0.82
100.03
170
40
94
36
76.7
0.15
11.3
2.85
0.07
0.02
0.17
0.01
2.72
5.48
–
0.74
100.21
190
55
180
34
76.4
0.13
11.1
1.85
0.19
0.01
0.23
0.55
3.46
5.0
0.05
0.82
99.79
170
43
180
35
3.7
15.2
5.02
630
13.3
75
16
27
79
8.6
35
8.7
0.2
9.1
1.55
10.2
2.23
6.07
0.91
5.72
0.84
0.07
3.4
1
77.0
0.1
11.9
1.65
0.15
0.01
0.29
0.48
3.9
4.08
–
0.56
100.12
140
13
70
29
3.62
15.4
4.14
600
13.2
78
20
29
69
8.7
35
8.8
0.14
8.8
1.53
9.8
2.18
6.01
0.9
5.65
0.83
0.05
3.7
0.91
76.8
0.12
11.7
1.76
0.22
0.01
0.18
0.37
3.6
4.24
–
0.86
99.86
140
21
55
32
78.4
0.1
10.1
1.98
0.11
0.02
0.05
0.48
2.65
5.3
–
0.43
99.62
180
38
200
31
77.3
0.1
11.3
2.7
77.2
0.09
10.85
1.49
0.19
0.01
0.39
0.48
3.2
4.81
–
1.48
100.19
170
67
120
33
74.6
0.24
12.4
1.18
0.36
0.05
0.25
0.12
2.78
6.36
–
0.98
99.32
150
30
300
34
6
71.4
0.36
12.8
3.0
0.2
0.05
0.48
1.30
3.62
4.45
0.06
1.56
99.29
95
70
250
38
6
69.8
0.44
13.2
3.34
0.37
0.03
1.45
0.98
3.71
4.66
0.08
1.91
99.97
90
85
300
29
2.9
49.53
2.02
15.1
8.11
3.32
0.14
4.46
4.7
5.65
1.15
0.49
4.74
99.41
15
380
520
17
400
210
240
47
43
30
26
35
34
70
8.2
33
7.4
0.6
7
1.3
7.6
1.6
4.7
0.67
4.4
0.69
0.25
5.5
0.92
36
65
8.5
35
7.7
0.6
7
1.3
7.1
1.5
4.5
0.64
4.3
0.7
0.25
6
0.84
35
80
9.2
37
7
0.56
5.7
0.95
5.6
1.1
3.3
0.53
4.4
0.46
0.27
5.7
0.84
70.2
0.39
14.55
2.74
0.11
0.04
0.61
0.44
5.19
4.08
0.10
1.08
99.53
96
96
840
16
2.23
10.5
3.49
360
7.8
31
17
17
66
4.7
20
5
0.87
4.9
0.84
5.1
1.13
3.1
0.46
3.03
0.47
0.54
4.1
0.89
57.3
1.09
17.6
2.65
3.47
0.09
3.15
2.09
5.50
3.58
0.29
2.76
99.57
97
470
780
3
410
73.3
0.2
12.95
2.47
0.15
0.06
0.3
0.38
5.11
4.0
0.05
0.50
99.47
84
25
760
18
1.94
9.4
3.52
390
7.7
37
17
32
88
8.4
32
6.2
0.82
5.2
0.83
5.2
1.17
3.18
0.49
3.08
0.49
0.44
7.4
0.98
600
570
70
68
0.98
0.92
76.6
0.12
11.4
1.8
0.16
0.01 0.01
0.2
0.22
0.3
0.1
3
3.6
5
4.65
0.05 –
0.62
99.96 99.28
150 220
40
49
190 210
27
32
6.6
17
550
560
76
64
55
1
30
85
7.2
28
7.8
0.09
8.2
1.5
9.4
1.9
5.9
0.95
5.9
0.79
0.03
3.7
0.92
0.9
550
640
70
75
0.96 0.96
(1) Samples 1–10 = quartz porphyry, 11–12 = feldspathic rhyolite porphyry, 13 = trachyandesite, 14 = trachydolerite.
(2) Eu/Eu⁎ = Eun/(Smn⁎Gdn)1/2; (Na + K) / Al, agpaitic index = (Na2O + K2O)/Al2O3, mol.%.
(3) Here and in Tables 2 and 3 hyphen = concentration of element below the detection limit.
(4) Sampling localities are shown in Fig. 1; sample A255 from Amram area.
22
76
28
Table 2
Chemical composition of representative rocks from composite dykes (CD) of the bimodal suite (wt.%, ppm)
Sample no.
A2606
# of CD in Fig. 1 CD-1
Lan/Ybn
(Na + K) / Al
A202
A198 D218a
CD-2
CD-4
CD-5
Quartz porphyry from interiors
68.61
71.7
76.2
73.8
0.46
0.33
0.2
0.29
13.55
13
12.05 12.4
2.67
0.76
1.78
2.65
0.92
1.81
0.16
0.26
0.05
0.04
0.01
0.02
1.57
0.56
0.13
0.2
1.11
0.85
0.26
0.44
5.46
2.75
3.85
3.86
4.04
6.34
4.92
5.11
0.15
0.06
–
–
1.33
1.9
0.58
0.44
99.92
100.1
100.14 99.8
100
145
140
115
300
110
72
81
670
460
280
240
10
25
20
30
4.71
11.5
3.69
190
430
270
470
11.3
19
37
42
63
17
56
42
117
94
15
11.3
58
47
10.4
10.3
0.64
0.73
10.7
9.5
1.34
1.58
7
8.8
1.35
1.77
4.18
4.99
0.65
0.74
4.15
4.93
0.64
0.76
0.19
0.23
9.7
6.1
0.98
0.88
0.97
0.96
D231
CD-6
A127
A1271
CD-7
73.4
0.28
12.35
2.22
0.46
0.04
0.42
0.78
3.25
5.89
–
1.19
100.28
130
52
450
28
72.3
0.28
12.6
4.22
0.23
0.04
0.58
0.4
3.36
5.35
–
1.26
100.62
120
52
320
29
13
11
390
440
37
21
59
110
53
20
48
95
58
43
0.95
0.9
74.1
0.25
11.6
2.99
0.34
0.03
0.92
0.49
3.55
4.0
–
1.64
99.91
91
41
60
45
3.54
12.7
3.92
620
14.3
74
28
61
132
15.1
59
11.9
0.66
11.4
2
11.2
2.24
6.38
0.94
6.11
0.92
0.17
7.2
0.88
A30
A663
A18
CD-8
73.2
0.35
11.8
2.66
0.72
0.03
0.79
0.53
3.34
4.77
0.05
1.26
99.5
120
56
70
51
4.15
11.6
3.76
660
14.1
75
23
60
131
15.9
63
12.4
0.72
11
1.76
10.7
2.3
6.19
0.93
5.81
0.87
0.19
7.4
0.9
76
0.16
11.6
2.79
0.32
–
0.05
0.48
4.6
3.0
–
0.83
99.83
95
30
50
32
4.13
17.5
4.46
590
14.1
64
23
28
65
7.2
32
9
0.2
8.5
1.7
10.3
2.2
6.9
1
6.7
1
0.07
3
0.93
76.2
0.13
11.6
2.74
0.18
0.01
0.13
0.19
3.0
5.4
–
0.82
100.4
194
60
150
34
2.75
17.7
5.75
590
14.4
75
19
43
102
12.2
50
11.6
0.28
11
1.93
11.2
2.33
6.73
1.02
6.57
0.98
0.08
4.7
0.93
74.7
0.14
11.2
2.6
0.16
–
–
1.13
3.72
3.88
–
2.12
99.75
135
50
50
35
3.01
17.3
4.74
590
14
68
21
60
120
14
55
11.5
0.3
11
1.9
11
2.3
6.9
0.98
6.5
0.98
0.08
6.6
0.92
A285
A2601
CDAmram
CD-1
74.6
0.2
11.8
2.12
0.13
0.02
0.34
0.24
2.88
5.84
–
1.33
99.5
150
56
240
30
16
410
44
20
58
106
43
A2603
A2604
A1961
N-1
A1978
A1977
CD-2
Mafic rocks from marginal zones
56.6
62.2
66.7
50.8
1.14
0.82
0.61
1.83
16.9
15.4
14
15.3
4.62
3.63
2.36
3.44
2.69
1.96
1.78
4.35
0.07
0.06
0.06
0.17
2.65
2.22
1.49
4.47
2.35
1.68
1.39
6.28
5.51
5
4.65
3.79
4.61
4.74
5.18
3.2
0.33
0.23
0.2
0.38
2.65
2.21
1.64
5.44
100.12 100.15 100.06 99.45
107
100
120
86
360
390
320
460
800
900
1130
940
7
9
11
13
1.17
4.2
1.15
225
180
190
210
5.2
24
21
19
28
23
24
52
7.1
30
6.1
1.5
6.4
0.8
4.2
0.8
2.35
0.33
1.96
0.29
0.74
8.8
A2022
A2021
CD-4
52.8
1.72
14.7
4.82
4.1
0.18
4.6
6.0
3.27
2.76
0.4
4.65
100
65
430
750
18
0.57
3.7
0.81
220
4.7
26
20
28
58
6.8
28
5.6
1.58
5.2
0.81
4.7
0.94
2.6
0.37
2.45
0.37
0.89
8.3
A1981
D21aa
A1272
A1411
CD-5
CD-6
CD-7
CD-8
50.47
2.77
14.75
11.56
–
0.1
4.37
3.2
3.32
3.36
1.19
5.05
100.14
102
167
657
22
1.47
4.5
1.49
318
6.8
48
34
34
82
11
50
11.0
3.19
10.1
1.53
7.9
1.5
3.95
0.53
3.37
0.48
0.92
7.3
55.0
2.17
14.0
9.9
1.17
0.08
3.77
2.89
4.28
2.07
0.52
4.37
100.22
52
140
430
26
1.01
5.8
1.91
350
8.2
51
27
40
90
11.3
48
9.9
2.19
8.9
1.35
8
1.68
4.26
0.62
3.86
0.57
0.71
7.5
65.4
1.03
13.8
5.1
0.53
0.05
2.32
1.73
3.58
3.72
0.25
2.2
99.71
130
130
460
23
60.9
1.36
14.25
6.44
0.78
0.08
2.65
2.1
4.16
3.25
0.35
3.46
99.78
92
140
640
23
420
390
51
48
61.3
1.04
13.9
2.02
3.14
0.11
2.5
3.82
3.5
4.44
0.2
4.23
100.2
92
310
490
19
65.5
0.67
13.3
1.65
2.12
0.08
1.1
3.04
2.51
5.97
0.12
3.41
99.47
140
190
690
21
61.7
1.15
14.65
3.08
2.25
0.07
2.7
1.93
3.47
5.26
0.39
2.84
99.49
150
200
740
14
64.1
1.06
14.65
2.06
2.09
0.04
2.27
2.0
4.3
4.19
0.33
2.53
99.62
115
160
460
13
49.3
2.88
14.7
6.57
4.35
0.14
4.99
4.3
3.83
3.22
1.07
5.02
100.37
73
400
840
14
320
370
360
370
280
35
35
54
52
45
37
84
11.1
44
8.7
1.18
9.0
1.17
6.2
1.18
3.48
0.52
3.26
0.51
0.41
8.1
43
94
12.4
48
9.2
0.92
9.6
1.2
6.6
1.26
3.77
0.58
3.65
0.56
0.3
8.4
D621
Y. Katzir et al. / Lithos 97 (2007) 336–364
SiO2
TiO2
Al2O3
Fe2O3
FeO
MnO
MgO
CaO
Na2O
K2O
P2O5
LOI
Total
Rb
Sr
Ba
Nb
Ta
Th
U
Zr
Hf
Y
Ga
La
Ce
Pr
Nd
Sm
Eu
Gd
Tb
Dy
Ho
Er
Tm
Yb
Lu
Eu/Eu⁎
A1971
0.94
Eu/Eu⁎ = Eun / (Smn ⁎ Gdn)1/2; (Na + K) / Al, agpaitic index = (Na2O + K2O)/Al2O3, mol.%.
343
344
Y. Katzir et al. / Lithos 97 (2007) 336–364
2000 times with 2% HNO3 before ICP-MS analysis. The
internal standard was used for monitoring the signal
shift during the measurement. The analytical precision
was about ± 3% (2σ). Sr and Nd isotopic compositions
were analyzed on unspiked samples. Powdered samples
were dissolved in Savilex bombs and followed by a
series of standard procedures until samples were
completely dissolved. Isolation of Sr and Nd were
achieved using a 2-column technique, and Sr fractions
were occasionally further purified using a third column.
The first column was packed with 2.5 ml cation
exchange resin (Bio-Rad AG50W-X8, 100–200 mesh)
and was used to collect Sr and REE fractions. The
second column used for Sr purification was packed with
1 ml cation exchange resin, identical to the above. The
second column used for Nd isolation was packed with
1 ml Ln-B25-A (Eichron) resin, which was covered on
top by a thin layer of anion exchange resin (Bio-Rad
AG1-X8, 200–400 mesh). Mass analyses were performed using a 7-collector Finnigan MAT-262 mass
spectrometer in dynamic mode. 143 Nd/144Nd ratios were
normalized against the value of 146Nd/144Nd = 0.7219,
whereas 87Sr/86Sr ratios were normalized to 86Sr/88Sr =
0.1194. During the course of analyses, measurements
on NBS-987 Sr standard yielded 87Sr/86Sr = 0.710320 ±
0.000036 (n = 50) using static mode); and =0.710237 ±
0.000020 (n = 8) using dynamic mode. The 87Sr/86Sr
ratios reported herein (Table 5) have been duly adjusted
to NBS-987 = 0.710250. Our measurements on La Jolla
Nd standard yielded 0.511864 ± 0.000006 (n = 4), and on
JMC Nd standard, 0.511821 ± 0.000016 (n = 6). The onestage model age (TDM − 1) is calculated assuming a linear
Nd isotopic growth of the depleted mantle reservoir from
εNd(T) = 0 at 4.56 Ga to +10 at the present time.
TDM −1 ¼ 1=klnf1 þ ½ð143 Nd=144 NdÞs−0:51315
=½ð147 Sm=144 NdÞs−0:2137g;
where s = sample, λ = decay constant of 147 Sm
(0.00654 Ga− 1).
The two-stage model age (TDM − 2) is obtained
assuming that the protolith of the granitic magmas has a
Sm/Nd ratio (or fSm/Nd value) of the average continental
crust (Keto and Jacobsen, 1987)
TDM −2 ¼ TDM1 −ðTDM1 −tÞðfcc−fsÞ=ðfcc−fDM Þ;
where fcc, fs, fDM =fSm/Nd are values of the average
continental crust, the sample and the depleted mantle,
respectively. In our calculation, fcc = − 0.4 and
fDM = 0.08592 are used, and t= the intrusive age of granite.
4. Petrography
Quartz porphyry and subordinate feldspathic rhyolite porphyry are grayish pink and brick red homogeneous porphyritic rocks, with very fine-grained and
Table 3
Composition of rocks collected in the cross section through the Netafim composite dyke # 2 (wt.%, ppm)
Sample no. A196-1 N-1
A196-2 A197-8 A197-7 N-6
N-2
A197 A197-1 A197-4 A197-5 N-7/1
A196-3 A197-6
Rock
QP
QP
TDc
TA
TA
TA
TD
SiO2
TiO2
Al2O3
Fe2O3
FeO
MnO
MgO
CaO
Na2O
K2O
P205
LOI
Total
Rb
Sr
Ba
Nb
Zr
Y
TD
TD
TA
TA
Tdc
Tdc
QP
1
2
3
4
5
6
7
8
9
10
11
12
13
14
50.8
1.83
15.3
3.44
4.35
0.17
4.47
6.28
3.79
3.78
0.38
5.44
100.03
86
460
940
13
210
28
52.8
1.72
14.7
4.82
4.1
0.18
4.6
6.0
3.27
2.76
0.4
4.65
100
65
430
750
18
220
26
55.2
1.54
14.6
3.27
4.01
0.25
4.8
5.35
3.76
3.09
0.33
4.04
100.24
72
410
670
15
240
27
61.3
1.04
13.9
2.02
3.14
0.11
2.5
3.82
3.5
4.44
0.2
4.23
100.2
92
310
490
19
320
35
65.5
0.67
13.3
1.65
2.12
0.08
1.1
3.04
2.51
5.97
0.12
3.41
99.47
140
190
690
21
370
35
67.2
0.7
13.0
2.82
1.41
0.03
1.69
1.65
3.21
5.66
0.15
2.24
99.76
140
170
370
32
380
34
71.7
0.3
12.7
1.46
1.23
0.02
0.53
1.19
2.38
6.6
0.06
1.59
99.76
160
150
530
37
440
43
72.1
0.32
13.3
1.64
1.02
0.04
0.76
0.55
3.71
5.19
–
1.39
99.99
105
105
240
26
430
43
71.7
0.33
13
0.76
1.81
0.04
0.56
0.85
2.75
6.34
0.06
1.9
100.1
145
110
460
25
430
37
66.4
0.63
13.2
1.67
2.12
0.08
1.55
2.73
2.96
5.65
0.11
3.17
100.27
135
230
500
22
380
34
61.5
1.09
14.1
1.33
3.74
0.11
3.15
3.24
3.51
4.61
0.22
3.41
100.01
100
290
540
21
310
33
58.2
1.26
13.95
3.87
2.94
0.1
3.62
3.52
3.62
4.3
0.26
3.94
99.58
98
340
590
20
270
25
59.7
1.23
14.5
1.84
4.1
0.11
3.85
3.88
3.46
3.5
0.24
3.61
100.02
69
330
420
19
280
37
49.6
2.15
15.6
3.89
5.93
0.22
6.4
7.08
3.74
1.73
0.43
3.52
100.29
37
560
560
12
200
23
Abbreviations: TD = trachydolerite; TA = trachyandesite; Tdc = trachydacite; QP = quartz porphyry.
Y. Katzir et al. / Lithos 97 (2007) 336–364
microcrystalline matrix. Phenocrysts of quartz, K-rich
alkali feldspar and rare albite constitute up to 10 vol.%;
in rhyolite porphyry quartz phenocrysts are absent. In
rocks that contain mafic schlieren and microgranular
enclaves some alkali feldspar phenocrysts are zoned:
thin brown rim enriched in albitic component is clearly
distinguished (Supplementary data, Table 1S). The matrix
345
of microgranophyric, cryptocrystalline and spherulitic
texture is made up mainly of the same minerals as the
phenocrysts; flakes of biotite and needles of iron oxides
are evenly scattered. Accessories are zircon and apatite.
The felsic dyke rocks are noticeably altered: biotite is
almost completely replaced by hematite and chlorite,
alkali feldspar phenocrysts are turbid, reddish brown, and
Fig. 3. Silica variation diagrams of selected major and trace elements in quartz porphyry dykes from the bimodal suite. Stippled bands designate the
supposed border zone between “pure” and contaminated rocks.
346
Y. Katzir et al. / Lithos 97 (2007) 336–364
spotty, and in the matrix alkali feldspar is partly replaced
by clay minerals.
Trachydolerite is a dark gray, very fine-grained rock
of porphyritic texture. Phenocrysts (3 to 5 vol.%) are
elongated prisms of unzoned plagioclase (An52–54,
Table 1S) and subhedral crystals of augite enriched in
Mg and Ca (Table 2S). The matrix is intragranular:
wedge-shaped interstices between plagioclase laths
(An41) are occupied by augite and Fe–Ti oxides.
Interstitial alkali feldspar is a subordinate mineral
(Table 1S). Other interstices are filled with hypocrystalline aggregates of chlorite. Apatite is a common
accessory mineral. The rock is considerably affected
by alteration as indicated by the abundance of secondary
chlorite, actinolite, carbonate, epidote and albite and by
extensive sericitization of plagioclase.
Trachyandesite and trachydacite commonly bear
evidence of hybrid origin. The hybrid rocks were studied
in detail in the Netafim dyke #2 that shows gradational
contacts between the felsic interior and mafic margins
(‘Geological setting’). The reddish gray trachydacite
adjacent to the felsic interior has textural features and
phenocryst population characteristic of quartz porphyry:
euhedral quartz and alkali feldspar including grains with
brown peripheral zones (Table 1S, samples A197-7 and
A197-4). However, rare elongated plagioclase prisms,
similar in shape to those in the trachydolerite, also occur.
In the matrix, along with quartz and K-rich alkali feldspar,
tiny plagioclase laths are observed.
The trachyandesite from the transitional zone is texturally similar to the adjacent trachydolerite, but along
with andesine and augite phenocrysts (Tables 1S and 2S),
xenocrysts of quartz and alkali feldspar are present. Alkali
feldspars mostly occur as shapeless, heavily corroded
relics. Quartz is commonly rimmed by pyroxene, which is
compositionally distinct from magmatic augite (Table 2S,
sample 197-5, phenocryst and rim). The matrix contains
plagioclase (An28) and augite, with irregular grains of
quartz and alkali feldspar. Their amount increases
systematically towards the trachydacite subzone.
Trachyandesite and trachydacite making up marginal
zones in composite dikes with abrupt contacts are often
heterogenous. This is exhibited by the irregular variations of mafic and felsic mineral proportions at a
distance of b 1 m and by the occurrence of xenocrysts.
Less common is homogenous trachyandesite that is
similar in appearance and texture to trachydolerite.
(REE). Data for representative samples are given in
Tables 1, 2, and 3, and sampling localities are shown in
Fig. 1. A full set of chemical data with sample localities
is given in the online version of the Supplementary data
(Table 3S, Fig. 2S).
Quartz porphyry. The chemical composition of
quartz porphyry varies widely both in simple (noncomposite) and composite dykes. In particular, silica
content ranges from 68.6 to 78.4 wt.% (Fig. 3, Table 1).
However, in the majority of samples from simple dykes
the range is narrow extending from 76 to 77.5 wt.%
SiO2. By contrast, quartz porphyry making up the
interiors of composite dykes is often less silica rich,
with SiO2 ≤ 74 wt.%. Decreased SiO2 content in felsic
rocks is characteristic of varieties with clear textural and
mineralogical evidence of contamination by mafic
material (mafic enclaves and schlieren, plagioclase
xenocrysts, alkali feldspar phenocrysts with brownish
outer zones enriched in albite molecule). Contamination explains not only the SiO2-depletion, but also the
enrichments in MgO, FeO, CaO and related trace
elements (Fig. 3) and the less pronounced negative Eu
5. Geochemistry
Fig. 4. Chondrite-normalized REE patterns for quartz porphyries from
simple (non-composite) dykes (asterisks) and from composite dykes
(CD), stippled area. A, “pure” quartz porphyry; B, contaminated quartz
porphyry. Here and in Figs. 6 and 11 chondrite values are from Sun and
McDonough (1989).
The geochemical data set contains 70 analyses. Of
these, 26 samples were analyzed for rare earth elements
Y. Katzir et al. / Lithos 97 (2007) 336–364
anomaly (Fig. 4A and B). In simple quartz porphyry
dykes visible evidence of contamination is rare.
However, the compositional similarity between homogenous silica-depleted quartz porphyry from simple
dykes and undoubtedly contaminated rocks from
composite dykes suggests that these dykes also
crystallized from hybridized rhyolite magma. In Fig. 3
the border between “pure” and contaminated quartz
porphyries is shown by convention at 74–75 wt.%
SiO2.
Uncontaminated (‘pure’) quartz porphyry corresponds chemically to alkaline granite (Fig. 5A and B)
with agpaitic index of [Na+K]/AlN0.9, and K2O (4.2–
5.4 wt.%) dominating over Na2O (Table 1). In the
standard classification diagram (Marian and Piccoli,
1989) rock compositions occupy a small field straddling
the border between metaluminous and peraluminous
granites (not shown). Quartz porphyry exhibits geochemical traits of A-type granite (Fig. 5C) formed in a
347
within-plate environment (Fig. 5D). Major and trace
element contents vary within narrow limits: the total
content of MgO, CaO and TiO2 is less than 1 wt.% and
FeOt = 1.5–2.6 wt.% (Fig. 3). Average Zr and Y contents
are fairly high, 570 and 69 ppm, respectively, whereas
Sr content is low, 13–72 ppm (Tables 1 and 2, Figs. 3
and 5C). ΣREE averages 180 ppm, chondrite-normalized REE patterns are flat (Lan/Ybn is about 3) and
negative Eu anomaly is well pronounced, with Eu/Eu⁎
value of 0.05–0.07 (Tables 1 and 2; Fig. 4A).
Trachydolerite contains 47–53 wt.% SiO2. It is
low-Al (Al 2 O 3 = 14.7–15.3 wt.%) and low-Mg
(Mg# = 40–55) with decreased CaO (3.2–6.3 wt.%).
It is characterized by high contents of Ba = 660 to 940
ppm, Zr = 210 to 318 ppm, and Y = 26 to 47 ppm
(Table 2). Trachydolerites are classified according
to their TiO2 content to moderate-Ti and high-Ti
varieties, with 1.7–1.8 and 2.8–2.9 wt.% TiO2, respectively. The two groups are similar in many
Fig. 5. Quartz porphyry compositions plotted in classification diagrams. A, 100⁎[(MgO + FeOt + TiO2) / SiO2] vs (Al2O3 + CaO) / (FeOt + Na2O + K2O)
diagram, after Sylvester (1989); H.F. C–A= highly fractionated calc–alkaline. B, SiO2 vs Agpaitic Index [(Na+K]/Al], after Liégeois and Black
(1987). C, Ga/Al vs Zr diagram; gray rectangles denote the fields of S-, I-, and M-type granitoids (after Whalen et al., 1987). D, Y + Nb vs Rb tectonic
classification diagram, after Pearce et al. (1984).
348
Y. Katzir et al. / Lithos 97 (2007) 336–364
line basalt. In the Zr–Ti–Y geotectonic classification
diagram (Fig. 6B) all data points plot in the field of
within-plate basalts.
Rocks of intermediate composition forming mafic
margins with abrupt contacts in composite dykes and
rare simple mafic dykes plot in the field of andesite and
dacite in the Nb/Y vs Zr/TiO2 classification diagram
(Fig. 6A). The chemical affinities of the intermediate
rocks are described below, in the discussion that concerns the interaction between coeval silicic and mafic
magmas.
6. Stable and radiogenic isotope ratios
6.1. Oxygen isotopes
Oxygen isotope ratios were measured in rockforming minerals from felsic and mafic rocks (quartz,
alkali feldspar, plagioclase, augite) and in zircon from
the quartz porphyry (Table 4; Fig. 7). The whole-rock
δ18O values that were measured in samples analyzed for
radiogenic isotopes are listed in Table 5.
In a trachydolerite from a composite dyke the δ18O
of clinopyroxene is 5.78‰ (Fig. 7), which is within the
δ18O range of clinopyroxene in the mantle (Mattey
et al., 1994a) and is similar to the δ18O(Cpx) of modern
basalts from the Taupo volcanic zone (New Zealand)
erupted through the continental crust (Macpherson et al.,
1998). This suggests that the oxygen isotope ratio of
Table 4
δ18O values (‰, SMOW) in minerals from dyke rocks
Fig. 6. Mafic rock compositions plotted in classification diagrams.
A, Nb / Y vs [(Zr/TiO2) ⁎ 0.0001] diagram, after Winchester and
Floyd (1977). B, Zr–Ti/100 − Y ⁎ 3 diagram, after Pearce and Cann
(1973). C, chondrite-normalized REE patterns for moderate-Ti and
high-Ti trachydolerite.
geochemical features, but the high-Ti rock is richer in
FeO t , HFSE (P, Zr, Y) and REE (Table 2). The REE
patterns of both subgroups are gently dipping and
almost parallel (Fig. 6C) with Lan/Ybn ≈ 8 and Eu/
Eu⁎ ≈ 0.85 (Table 2).
The high proportion of secondary minerals (chlorite,
actinolite, sericite, carbonate, hematite) even in the
least altered samples suggests that the observed
contents of alkalis and other mobile elements might
not correspond to the original magmatic values.
Therefore classification diagrams based on immobile
trace elements are used. In the Nb/Y vs Zr/TiO2
diagram (Fig. 6A) compositions of both moderate-Ti
and high-Ti trachydolerites plot in the field of subalka-
Sample no.
Rock
Bimodal dyke suite 3
A197-6
trachydolerite
A52
quartz porphyry, stock
A197
quartz porphyry, dyke
A197-1
″
A-129
″
A-157
″
A198
″
A202
″
A128
″
A149
″
A285
″
D218a
″
Qtz
9.10
8.57
8.60
8.9
8.71
9.69
9.65
8.75
8.9
8.3
8.68
Afs [Pl]
Zrn [Cpx]
[11.16]
11.86
13.27
13.12
[5.78]
6.49
14.73
12.95
13.41
10.67
14.1
15.45
13.43
Dyke suite 2
A148
andesite porphyry
A160-1
dacite porphyry
[11.13]
[13.58]
Dyke suite 1
A173a
microdiorite
A28
lamprophyre
A25
″
[11.12]
[10.06]
[10.84]
6.45
7.23
Y. Katzir et al. / Lithos 97 (2007) 336–364
349
Fig. 7. Oxygen isotope ratios of mineral separates from the Late Neoproterozoic dyke suites of southern Israel. The SiO2 (wt.%) content is denoted
in parentheses. The shaded band marks the range of δ18O values of clinopyroxene in mantle xenoliths (Mattey et al., 1994a,b). The dashed
rectangle illustrates the δ18O range of plagioclase in oceanic arc lavas (Eiler et al., 2000). Also shown are the average δ18O values of
clinopyroxene phenocrysts from basalts and andesites of the Taupo arc, New Zealand and of plagioclase from the Lanai volcano, Hawaii (Eiler
et al., 1996; Macpherson et al., 1998). The Zrn–Afs–Qtz bar shows equilibrium fractionations between the minerals at magmatic temperatures
(after Chiba et al., 1989; Valley et al., 2003).
clinopyroxene from the trachydolerite was not modified
after the initial isotope equilibrium with the mantle
source rock. Notwithstanding, plagioclase of the
trachydolerite has a non-magmatic high δ18O value of
about 11‰ (Fig. 7). Similar values are also characteristic of plagioclase in mafic rocks from earlier dyke
suites (Fig. 7). Experimentally calibrated plagioclaseclinopyroxene isotope fractionations at temperatures of
crystallization of basaltic melts (1100–1200 °C)
are ≤ 0.7‰ (Chiba et al., 1989). Oxygen isotope ratios
of plagioclase from recently erupted oceanic arc lavas
range from 5.6 to 5.9‰ (Eiler et al., 2000). Likewise
δ18O(Pl) of historic subaerial alkaline basalt from the
Lanai volcano, Hawaii is 6.08‰ (Eiler et al., 1996).
Thus, the δ18O values of plagioclase in the Elat mafic
dykes are 4 to 5‰ higher than their original magmatic
equilibrium values.
Oxygen isotope ratios of zircon in quartz porphyry
from composite dykes and from a subvolcanic stock
range from 6.45 to 7.23‰. These values are significantly higher than δ18O of mantle zircon: 5.3 ± 0.6‰
(2σ; Valley, 2003). Since zircon is the best preserver of
initial magmatic values (Valley et al., 1994; Page et al.,
2006) the relatively high δ18O(Zrn) of quartz porphyry
suggests contribution of upper crustal sources to the
rhyolite magma. This stands in contrast to the mantle
δ18O of clinopyroxene from the trachydolerite and
suggests variable sources for the coexisting magmas.
In 11 samples collected from different quartz
porphyry dykes, the δ18O values of quartz range from
8.30 to 9.70‰ (Table 4; Fig. 7). The measured Δ18O
(Qtz–Zrn), 2.15 to 2.61‰, does not differ markedly
from the predicted magmatic quartz–zircon isotope
fractionation, 2.3‰ at T = 800 °C (Valley et al., 2003).
350
Table 5
Nd–Sr and δ18O (WR) isotope compositions of rocks from the dyke suite 3 (southern Israel) and the Katharina ring complex (Sinai, Egypt)
Sample no.
Rock
Katharina ring complex (590 Ma)
D163
syenogranite, pluton
D165
"
IL-66
quartz monz. porphyry
IL-161
quartz syenite porphyry
AG-32
peralk. granite porphyry
87
86
86
Rb/
Sr
Sr/
Sr
Sr
(ppm)
144.6
211.5
102.7
97.8
119.9
86.3
106.6
70.0
40.7
101.8
78.3
103.8
50.9
73.7
126.0
54.9
157.7
457.1
698.7
166.7
5.38
5.94
5.88
3.86
2.76
4.58
1.96
0.44
0.17
1.77
0.780104
0.781755
0.785968
0.759867
0.742676
0.773803
0.728985
0.708304
0.707539
0.727299
220.5
238.3
53.8
99.6
204.9
91.29
86.75
396.9
305.3
8.96
7.01
7.98
0.37
0.94
69.85
0.758037
0.767362
0.706085
0.710773
1.291379
±2σ
εNd
(0)
εNd
(T)
f
(Sm/Nd)
TDM − 1
(Ma)
TDM − 2
(Ma)
δ18O
WR
0.512484
0.512544
0.512606
0.512510
0.512465
0.512476
0.512553
0.512502
0.512486
0.512494
0.512478
0.512535
9
8
10
12
18
15
10
10
14
11
12
12
− 3.0
− 1.8
− 0.6
− 2.5
− 3.4
− 3.2
− 1.7
− 2.7
− 3.0
− 2.8
− 3.1
− 2.0
2.3
2.4
0.9
3.1
1.4
2.5
1.4
2.7
2.6
2.0
2.3
3.3
−0.36
−0.28
−0.10
−0.38
−0.32
−0.38
−0.20
−0.36
−0.37
−0.32
−0.37
−0.36
1154
1267
1077
1001
1063
1304
1111
1591
1128
1115
1242
1145
1070
1030
1118
1083
1008
11.71
12.76
11.49
12.16
11.52
11.15
9.79
10.54
9.24
10.5
10.92
13.1
0.512519
0.512546
0.512571
0.512503
0.512558
3
3
3
3
3
− 2.3
− 1.8
− 1.3
− 2.6
− 1.6
3.6
3.9
5.5
4.6
4.5
−0.40
−0.39
−0.46
−0.49
−0.41
1002
991
822
870
923
1010
971
911
1012
945
I
(Sr)
Sm
(ppm)
Nd
(ppm)
147
143
144
144
15
11
14
16
11
17
13
10
14
12
0.73483
0.73180
0.73648
0.72740
0.71943
0.73529
0.71253
0.70458
0.70612
0.71241
11.51
11.59
9.41
11.87
10.33
8.14
5.84
6.64
5.63
11.03
6.73
9.87
55.33
49.75
32.18
58.84
46.70
40.59
22.54
31.82
27.68
50.04
32.73
47.30
0.1258
0.1408
0.1768
0.1220
0.1337
0.1212
0.1566
0.1262
0.1230
0.1333
0.1243
0.1262
8
8
7
7
10
0.69904
0.70020
0.70300
0.70284
0.70372
6.34
7.26
10.97
12.08
12.17
32.57
36.33
62.42
72.82
63.49
0.1177
0.1208
0.1062
0.1003
0.1159
±2σ
Sm/
Nd
Nd/
Nd
(1) To obtain meaningful model ages, TDM were calculated for samples with f(Sm/Nd) that range from − 0.2 to − 0.6 (Wu et al., 2002).
(2) For mafic rocks only one-stage model age (TDM-1) was calculated.
(3) Samples IL-66, IL-161 and AG-32 are from ring dykes.
9.18
8.15
7.12
7.13
9.19
Y. Katzir et al. / Lithos 97 (2007) 336–364
Dyke suite 3 (590 Ma)
A18
quartz porphyry
A66-3
"
A30
"
A127
"
A198
"
A147-2
rhyolite porphyry
A156-1
"
A196-1
trachydolerite
N-1
"
D21aa
"
A196-4a
trachyandesite
A127-2
"
87
Rb
(ppm)
Y. Katzir et al. / Lithos 97 (2007) 336–364
This attests that in most dykes quartz was subjected to
only minor modification of the magmatic δ18O during
post-crystallization cooling.
In contrast to the rather limited variation in oxygen
isotope ratios of quartz (∼ 1.4‰, n = 10), the δ18O values
of alkali feldspar range from 10.67 to 15.45‰. The δ18O
(Qtz–Afs) values are always negative (reversed) and
range mostly from − 4 to − 6‰ (Table 5; Fig. 13), which
is far from the experimentally determined magmatic
equilibrium fractionation of +1‰ at 700 °C (Clayton
et al., 1989). Thus, the δ18O values for alkali feldspar in
quartz porphyry were extremely elevated by a postmagmatic process.
Whole-rock samples from felsic and mafic rocks
(Table 5) are characterized by high δ 18 O values,
commonly N10‰, mostly affected by the elevated
δ18O values of feldspars, which are the most modally
abundant phase in both groups and most readily
exchange oxygen during sub-solidus alteration. The
whole-rock oxygen isotope data are discussed in the
next section.
351
Fig. 8. εNd(590 Ma) vs (87Sr/86Sr)0 in dykes from the bimodal suite.
Note the narrow range of εNd(T) values in mafic and felsic rocks
irrespective of (87Sr/86Sr)0 values.
similar to that of the Katharina Complex. Comparatively
young model ages are characteristic of most Precambrian rocks throughout the Arabian–Nubian shield
supporting the idea of juvenile Neoproterozoic continental crust in the shield (Stein and Goldstein, 1996).
6.2. Radiogenic isotope ratios
7. Discussion
Rb–Sr and Sm–Nd data for 12 whole-rock samples
including silicic and mafic dyke rocks are given in
Table 5. For comparison, isotope data for the Katharina
ring complex (southern Sinai), which is made up of Late
Pan-African A-type granite and related ring dyke rocks
(Eyal and Hezkiyahu, 1980; Katzir et al., 2007), are also
presented in Table 5.
The calculated initial 87Sr/86Sr ratios (= I(Sr) at
590 Ma) range from ca. 0.704 (trachydolerite A196-1)
to ca. 0.737 (quartz porphyry, A-30). The I(Sr) values in
felsic rocks are not only high, but also extremely
variable, ranging from 0.71253 to 0.73648 (Table 5),
which precludes calculation of an isochron and suggests
postmagmatic disturbance of the Rb–Sr system.
The Sm–Nd isotopic data seem to provide much
more coherent information about the petrogenesis of the
dyke rocks. The initial 143 Nd/144Nd ratios, expressed as
εNd(T) values do not correlate with the initial Sr values
(Fig. 8). They are fairly uniform and range from + 0.9
to + 3.3. Given the analytical error of ± 0.5 ε-unit, there
is no systematic and significant difference between εNd
(T) values of felsic and mafic dyke rocks. However, the
range of εNd(T) values is clearly lower than that of the
Katharina Complex (+ 3.6 to +5.5). The difference is
also reflected by the oxygen isotope data (see Table 5,
last column) to be presented below. The Sm–Nd model
ages, calculated for felsic rocks with a two-stage model,
range from ca. 1000 to 1100 Ma. This range is rather
7.1. Effect of alteration on the original chemical
composition of the dyke rocks
The abundance of secondary minerals and the
significant elevation of the oxygen isotope ratios of
feldspar even in the seemingly least altered felsic and
mafic rocks indicate that the dyke rocks were subjected
to rather extensive alteration. Also the calculated wholerock (87Sr/86Sr)0 values for 590 Ma in quartz porphyry
and rhyolite porphyry are unusually high and vary over
a wide range. The Rb–Sr isotope data do not allow
calculating an isochron. This raises the question of how
well the original chemical composition of the dyke
rocks has been preserved under conditions of postmagmatic alteration. We approached this problem by
comparing quartz porphyry and trachydolerite from
southern Israel to similar rock types from adjacent
regions, the Northeastern Desert, Egypt and southwestern Jordan. According to Stern et al. (1988) and Jarrar
et al. (2004), these rocks have been subjected to rather
mild alteration that did not prevent dating them by the
Rb–Sr method. Unfortunately, there is not a single key
area in which both felsic and mafic rocks can be used for
comparison as reference rocks. In particular, in
southwestern Jordan the rhyolite from composite
dykes is chemically uniform and contains 75–77 wt.%
SiO2, but mafic margins are made up of hybrid latite
352
Y. Katzir et al. / Lithos 97 (2007) 336–364
(Jarrar et al., 2004). In the Northeastern Desert dolerite
is abundant in the bimodal dyke swarm, whereas
rhyolite is characterized by wide range of SiO2, from
69 wt.% to 77 wt.%, and the origin of low-SiO2 rhyolite
is debatable (Stern et al., 1988). For this reason we have
chosen the dolerite from the Northeastern Desert and
rhyolite from southwestern Jordan as reference rocks for
normalization. A comparison of normalized average
volatile-free compositions of felsic rocks with ≥ 75 wt.
% SiO2 and mafic rocks with 49–56 wt.% SiO2
(Table 2; supplementary data, Table 4S) is presented
by spidergrams (Fig. 9).
The patterns in Fig. 9A show that within one standard
deviation the felsic rocks from all three regions are fairly
similar in Th, Y, La, Sm, Gd and in the mobile LIL
element contents (Rb, K, and Na). Rhyolites from
southern Israel and the Northeastern Desert, as compared to felsic rocks from southwestern Jordan, are
richer in Nb, Zr, Ce, Yb and poorer in Sr. This points to
their more alkaline character and suggests that the
observed chemical distinctions are caused by original
compositional features of the silicic magmas in the
Fig. 9. Spidergrams of felsic (A) and mafic (B) rocks from the bimodal
dyke suite in southern Israel normalized to well-preserved Late
Neoproterozoic dyke rocks from similar suites in Egypt and Jordan
(see text for more details). A, normalized to rhyolite from southwestern
Jordan (Jarrar et al., 2004); B, normalized to dolerite from the
Northeastern Desert, Egypt (Stern et al., 1988). Stippled areas are ±1σ
range of rock compositions used for normalization. A full set of data is
given in Table 3S of the supplementary data.
different regions rather than by migration of elements in
the process of alteration. It is thus concluded that the
initial igneous composition of the felsic dyke rocks is
sufficiently well preserved.
Chemical data for mafic rocks are compared in Fig. 9B.
The plot shows that the chemical compositions of
trachydolerite from southern Israel and of mafic rocks
from the bimodal dyke suite in Egypt are essentially
similar. The main distinction is the higher K and Rb
contents of the S. Israel rocks. It is conceivable that the K
and Rb enrichment was partly caused by metasomatic
alteration (Agron and Bentor, 1981; Wachendorf et al.,
1985). The original concentration of these elements cannot
be quantitatively assessed. We assume that it is comparable to that of the mafic dykes from the Northeastern
Desert, namely about 2 wt.% K2O and ∼50 ppm Rb.
7.2. Interaction of mafic and silicic magma
Field, petrographic and chemical data indicate that
interaction between mafic and silicic magmas occurred
during the formation of the bimodal dyke suite.
Interaction occurred both in situ, in the process of
composite dyke formation, and at greater depth from
which hybrid magmas were injected to the subsurface
level.
The most graphic example of in situ interaction
between rhyolite and basalt magmas is the Netafim dyke
#2 (Fig. 10). Symmetric structure and gradational
contacts between felsic interior and mafic margins
show that the injection of silicic magma occurred when
part of the basic melt had not yet solidified. The
phenocryst assemblages in the trachydacite and trachyandesite from the transitional zone were derived
from both end-member magmas. This points to mutual
penetration of two adjacent partially crystallized
magmas that already contained phenocrysts.
Variation of rock chemistry across dyke #2 is shown
in Table 3 and in a set of diagrams (Fig. 10). Both major
and selected trace elements (except Ba) demonstrate
symmetrical and regular change from quartz porphyry
through the hybrid trachydacite and trachyandesite to
trachydolerite. The only noticeable discrepancy in the
distribution of Na2O, K2O, and Rb is seen in one sample
from quartz porphyry (sample 8) that contains small
mafic schlieren. Progressing from quartz porphyry to
trachydolerite, REE contents decrease and Eu/Eu⁎
increases systematically as expected in mixing of silicic
and mafic end members (Fig. 11).
In a series of silica variation diagrams (Fig. 12) the data
points are mostly arranged along straight regression lines
connecting end-members, with R-values close to unity.
Y. Katzir et al. / Lithos 97 (2007) 336–364
353
Fig. 10. Compositional profiles across the Netafim composite dyke #2 (plots of selected major and trace elements versus distance). Zones within the
dyke: TD = trachydolerite; mixing zone: TA = hybrid trachyandesite, Tdc = hybrid trachydacite; QP = quartz porphyry. Sample numbers are the same
as in Table 3. Major elements, wt.%; trace elements, ppm.
The scattering in the Ba and Rb data possibly stems from
fast chemical diffusion superimposed on bulk mixing.
Thus, chemical data are in a good consistency with
structural and petrographic evidence indicating that the
Fig. 11. Chondrite-normalized REE patterns for rocks from the
Netafim composite dyke #2. The pattern numbers correspond to the
sample numbers in Table 3; in parentheses: silica content (wt.%).
main process responsible for the formation of the hybrid
rocks in the composite dyke #2 was magma mixing.
Mass balance calculations performed for major and
trace elements show that the trachyandesite could form
by mixing of the basalt and rhyolite components in
proportions of 57% and 43% respectively, whereas for
trachydacite the proportions are 20.5% and 79.5%,
respectively (Table 6, calculated using program Igpet for
Windows, 1994).
Composite dykes with gradational contacts like
composite dyke #2 are not common in the Elat area.
Actually, dykes with abrupt contacts are dominant and
in many of them the mafic margins are made up of
trachyandesite and trachydacite. In places these rocks
make up simple dykes. The intermediate rocks are often
heterogeneous and contain xenocrysts of quartz, Kfeldspar and plagioclase with sieve-like texture. These
features show that magma mixing was important in the
genesis of the intermediate rocks. Abrupt contacts in
composite dykes attest that hybrid magmas were
produced mainly at greater depth prior to emplacement.
354
Y. Katzir et al. / Lithos 97 (2007) 336–364
Fig. 12. Silica variation diagrams for rocks making up the Netafim composite dyke #2 with regression lines and the correlation coefficient values (R).
Stippled fields denote compositions of intermediate rocks from composite dykes with abrupt contacts, and simple andesite dykes; I and II, groups of
dyke rocks (see text).
Y. Katzir et al. / Lithos 97 (2007) 336–364
355
Fig. 12 (continued ).
Chemical compositions of intermediate rocks are plotted
in binary diagrams (Fig. 12) and show rather restricted
similarity with the in situ produced hybrid rocks from
the Netafim dyke. Only less mobile major elements tend
to be arranged close to the mixing lines. Elements with
high diffusivity, K2O and Rb, demonstrate large scatter,
which points to the great role of diffusive exchange
between co-existing magmas (Zorpi et al., 1991; Poli
and Tommazini, 1991; Litvinovsky et al., 1995b). The
distinctions in other trace elements are more significant.
Intermediate rocks are clearly divided into two groups
with different Sr, Ba, Zr, Nb, and Y concentrations, and
neither of them is similar to the in situ formed hybrid
rocks (Fig. 12, Tables 1 and 2; Table 3S, supplementary
data). It is likely that magma mixing processes had not
been frozen at the initial stage as it occurred in the
Netafim dyke. At greater depth mixing could have
progressed to much higher extent and may have been
followed by fractional crystallization and assimilation.
Some variations in the end-member compositions are
also possible.
7.3. Oxygen isotope evidence of late alteration of dyke
rocks
Initial magmatic 18O/16O ratios have been retained
only in two minerals from the dyke rocks: zircon in
felsic rocks and clinopyroxene in trachydolerite. The
relatively small variation in δ18O(Qtz) and the magmatic-like Δ18O(Qtz–Zrn) in the dykes indicate that the
oxygen isotope ratio of quartz was only slightly
modified after crystallization (Fig. 7). However, the
magmatic δ18O values of feldspars were significantly
altered. Isotope ratios of quartz and alkali-feldspar from
quartz porphyry dykes are presented in a δ–δ diagram
(Fig. 13). For comparison, isotope data on alkaline
granites from southern Israel and Sinai Peninsula (the
Katharina ring complex) are also shown (the data are
compiled in Table 7). The δ18O values for quartz in each
rock unit is almost uniform, while δ18O(Afs) is highly
variable thus forming well-defined, narrow and vertically elongated arrays in the plot. A notable difference
between granites and dykes is clearly exhibited: Δ18O
(Qtz–Afs) in quartz porphyries are always negative and
range mostly from − 2 to − 6‰, whereas in the granites
these values vary from − 0.7 to − 2.0‰. Measured
quartz–alkali feldspar isotope fractionations in fresh
granites worldwide are typically about + 1 to + 1.5 ‰
(Criss and Taylor, 1983). These are slightly larger than
the experimentally determined equilibrium Δ18O(Qtz–
Ab) at 800 °C (0.8‰; Clayton et al., 1989) and probably
reflect inter-mineral subsolidus oxygen exchange during
cooling. It follows that the magmatic 18O/16O ratio of
feldspar was elevated by exchange with external
reservoir both in plutonic and dyke rocks, but in the
latter the elevation was extreme (up to 8‰).
Quartz is much more resistant than feldspar to
hydrothermal 18O/16O exchange. The preservation of
magmatic δ18O values in quartz on the one hand and the
marked increase in δ18O(Afs) on the other hand indicate
that dykes experienced intense water-rock interaction.
However, high-temperature (250–500 °C) hydrothermal
alteration by circulating meteoric water heated by a
cooling pluton, which is indicated in many shallow
intrusions (e.g., Criss and Taylor, 1983), can be ruled
out. Oxygen exchange with hot meteoric water would
instead appreciably decrease the isotopic ratio of
feldspar relative to quartz to form highly positive
quartz–alkali feldspar fractionations (Criss and Taylor,
1983, 1986). A simple closed-system, water-rock
exchange model indicates that the observed high δ18O
values of feldspar in the felsic dykes (+ 11 to +16‰)
356
Y. Katzir et al. / Lithos 97 (2007) 336–364
Table 6
Results of modelling of magma mixing for the Netafim composite dyke #2
Sample wt.%
SiO2
TiO2
Al2O3
FeOt
MgO
CaO
Na2O
K2O
P2O5
Sum residuals squared
Mixing components (wt.%)
Basic magma
Felsic magma
ppm
Rb
Ba
Sr
Zr
Nb
Y
La
Ce
Nd
Sm
Eu
Gd
Dy
Er
Yb
Mixing magmas
Hybrid magmas
Trachydolerite
Qtz porphyry
Trachyandesite (n = 5)
Trachydacite (n = 3)
n=3
n=3
Observed
Calculated
Observed
Calculated
54.28
2.02
16.15
8.97
5.48
6.85
3.83
[2.00]
0.42
73.15
0.33
13.24
2.56
0.63
0.88
3
6.15
0.05
61.79
1.28
14.83
6.07
3.74
4.14
3.73
4.17
0.26
62.07
1.28
14.81
6.16
3.36
4.25
3.45
3.77
0.26
0.40
68.69
0.69
13.63
3.85
1.50
2.55
2.99
5.96
0.13
68.99
0.67
13.78
3.86
1.62
2.09
3.16
5.28
0.13
0.74
56.8
43.2
[50]
750
483
210
14
26
26
55
29
5.9
1.54
5.8
4.5
2.48
2.21
137
410
122
433
29
41
56
117
58
10.4
0.64
10.7
7
4.18
4.15
86
543
336
285
19
32
37
84
44
8.7
1.18
9
6.2
3.48
3.26
87
600
326
305
21
32
39
81
41
7.8
1.1
8
5.5
3.2
3
20.5
79.5
138
520
197
377
25
34
43
94
48
9.2
0.92
9.6
6.6
3.77
3.65
118
478
195
386
26
38
50
104
52
9.4
0.8
9.7
6.4
3.89
3.7
(1) Modelling was done using the program IGPET (Terra Softa Inc., 1994).
(2) In rock compositions major elements are recalculated to total = 100% volatile free.
(3) In trachydolerite, accepted contents of K2O = 2 wt.% and Rb = 50 ppm are in brackets (see Section 7.1).
(4) Total Fe as FeOt.
require very low temperatures of alteration (≤100 °C),
if it occurred by exchange with typical meteoric water of
δ18O(H2O) = − 5 to − 15‰ (Wenner and Taylor, 1976).
Low-temperature alteration of quartz porphyry is also
indicated by extensive reddening and turbidization of
alkali feldspar. Likewise, plagioclase in the mafic dykes
is commonly sericitized, turbid and characterized by
significant loss of Ca, culminating in almost complete
albitization of labradorite–andesine (Litvinovsky et al.,
in press). Therefore the 4 to 5‰ elevation of δ18O (Pl)
relative to its inferred magmatic ratios (Fig. 7) could
also have occurred by low-temperature interaction with
meteoric water. Alternatively, higher temperature alteration should involve unusually 18O/16O-rich waters.
High temperature (∼ 250 °C) interaction with high-δ18O
brines (+ 3‰) was suggested to account for the
moderately negative Δ18O (Qtz–Pl) (− 1 to − 1.4‰)
measured in Timna mafic and felsic plutonic rocks
(Beyth et al., 1997). The hydrothermal activity presumably responsible for the 18O/16O enrichments in the
Timna igneous complex was attributed to circulation of
brines through fractures related to the Dead Sea
transform (Beyth et al., 1997; Matthews et al., 1999).
The alleged timing of the hydrothermal event was
largely based on the chemical remanent magnetization
of Miocene age measured in the Timna igneous rocks
(Marco et al., 1993). Here we show that postmagmatic
18
O/16O elevations in feldspars, and consequently in
whole rocks, are not restricted to the Timna area, but
characterize dykes and plutons throughout the order of
magnitude larger Elat area. Moreover, Late Neoproterozoic alkaline granite, the Katharina pluton, that
Y. Katzir et al. / Lithos 97 (2007) 336–364
357
the data points of the dyke rocks fall in the first quadrangle,
which is totally unexpected from felsic igneous rocks.
This indicates a pronounced disturbance of the pristine
Rb–Sr isotope system that could have occurred either at
the magmatic or post-crystallization stage.
Table 7
δ18O values (‰, SMOW) in minerals from plutonic and metamorphic
rocks from southern Israel and the Katharina ring complex, Sinai
Sample no.
Rock
Qtz
Afs [Pl]
Elat granitic gneiss, 740 Ma (Eyal et al., 1991)⁎
YE-8
11.81 10.79
YE-9
9.97
9.45
YE-16
11.44 10.54
YE-24
10.19 9.91
Fig. 13. δ–δ diagrams for quartz and alkali feldspar in quartz porphyry
from the bimodal suite (asterisks). Isotherm for equilibrium fractionation between Qtz and Ab at 800 °C is after Clayton et al. (1989). For
comparison, data for three Late Neoproterozoic granite complexes are
presented (see text).
outcropsN 50 km away from the Dead Sea transform in
central southern Sinai also shows mild 18 O/ 16 Oenrichments in its uppermost part that is comparable
to those measured in Timna. The origin(s) and timing(s)
of the elevation of isotope ratios that is characteristic of
a large area in the northernmost ANS are yet to be
determined. Our results show, however, that late oxygen
exchange with external fluids occurred not only in the
bimodal dyke suite, but also in dykes from the two
preceding suites (Fig. 7; Table 4). This attests that the
alteration process was not linked with injection of a
specific batch of magma. Presumably the extensive
water–rock exchange happened mainly in dykes since
dyke injections are commonly controlled or followed by
fault zones that accommodate circulation of water. Less
intense 18O/16O variations in the host granites (Fig. 13)
may be explained by the fact that these rocks were
sampled at sites with no evidence of fracturing and rock
alteration.
7.4. Origin of the unusual I(Sr) values in the quartz
porphyry dykes
The I(Sr) (590 Ma) values for individual quartz and
rhyolite porphyry dykes from the bimodal suite vary
widely, from 0.71253 to 0.73648 (Table 5), and the Rb–
Sr isotope data do not allow calculation of a realistic
isochron. In a classic diagram of I(Sr) vs εNd(T) (Fig. 8),
Calc-alkaline
YE-11
YE-33
YE-34
YE-30
Elat granite, 620 Ma (Eyal et al.,
Monzogranite
10.38
″
10.31
″
10.71
″
11.17
2004)
10.51
9.42
10.06
9.83
Zrn
6.67
7.09
6.68
7.23
7.15
6.97
6.36
Alkaline Yehoshafat granite, 605 ± 4 Ma (unpublished authors' data)
YE-36
syenogranite
9.26
11.18
6.76
YE-10
″
9.27
9.00
YE-35
″
9.22
10.35
6.52
AG-81
″
9.13
9.85
6.6
Timna complex, southern Israel, ∼ 610 Ma (Beyth et al., 1994)
AG-40
Quartz syenite
8.95 [7.72] 5.75
AG-62
″
8.68 [7.94] 5.82
AG-42
Alkali–feldspar granite 8.17
10.28
AG-63
″
8.5
8.52
5.5
AG-63a
″
5.55
AG-67
″
8.67
8.11
5.48
AG-68
″
8.4
10.53
Katharina ring complex, Sinai Peninsula, ∼ 590 Ma (Katzir et al.,
2007)
D111
Syenogranite, pluton
8.48
8.75
D112
″
8.03
7.89 [7.86]
D130
″
8.55
8.33 [7.58]
D156
″
8.18
7.59 [7.75]
Avg (n = 6) ″
5.82 ±
0.06
D100
Ab–Afs granite, pluton 8.47
9.54 [8.31]
D59
″
8.34
9.64 [9.12]
D79
″
8.21
9.14
D45
Perthite granite, pluton 8.17
9.19
D46
″
8.27
9.66
D48
″
8.15
9.17
IL-8
Perthite quartz syenite– 8.22
8.65
5.48
IL-84
–porphyry, ring dyke
7.79
8.35
5.66
IL-112
″
7.87
IL-147
″
7.91
5.49
IL-9
″
8.32
5.50
(1) ⁎ = here and below in the table, sources for age data are given.
(2) δ18O values from Yehoshafat and Timna rocks after Gal, 2006.
(3) Avg (n = 6) = composite separate of zircon from 6 samples.
358
Y. Katzir et al. / Lithos 97 (2007) 336–364
At the magmatic stage, such a disturbance could
have been caused by active interaction between coeval
silicic and mafic magmas. Some experimental studies
have observed that isotopic gradients homogenize
roughly ten times faster than their respective chemical
gradients, so a basaltic magma may thereby impart its
isotopic characteristics to a silicic magma long before
the chemical composition of the latter is noticeably
changed (Baker, 1989, 1990; Lasher, 1990, 1994;
Snyder and Tait, 1998). Also it was shown that the
87
Sr/86Sr of silicic magmas is very sensitive to input
from basaltic injections, but the Nd-isotope ratios are
preserved (Snyder and Tait, 1998).
Applying the isotope exchange model to the bimodal
dykes of Elat, we assume that the mafic end-member
had I(Sr) of about 0.7050 and εNd(T) = 2, whereas the
felsic end-member, i.e. the least hybridized silicic
magma, had I(Sr) ∼ 0.7400 and εNd(T) ∼ 2 (Table 5).
Such isotope characteristics of the supposed silicic endmember are unusual. On the one hand, very high I(Sr)
value points to typical ancient continental crust source;
on the other hand, the εNd(T) values are characteristic of
mildly depleted mantle or of juvenile crust. However,
the main argument that testifies against isotope
exchange during magma mixing is the absence of felsic
rocks with high I(Sr) values in the territory of the ANS.
All granites and gneisses that may be regarded as
possible silicic end-member sources have I(Sr) values
of b 0.710–0.715, which are characteristic of crust in
juvenile provinces like the ANS (Stein and Goldstein,
1996; Stein, 2003).
In order to assess the role of post-crystallization
processes in disturbing the Rb–Sr system, one has to
acknowledge the crucial effect of alkali feldspar, the
major carrier of both Rb and Sr in the felsic dykes from
the bimodal suite. The water–rock interaction that
resulted in the elevation of δ18 O(Afs) up to 13–15‰
could also cause the significant disturbance of the Rb–
Sr isotope system. Indeed, a fairly clear positive
correlation between ( 87 Sr/ 86 Sr)0 and δ 18 O(WR) in
rhyolitic rocks is demonstrated in Fig. 14A. By contrast,
in the Katharina ring complex disturbance of the initial
oxygen isotope is minimal (Fig. 13), and the
(87 Sr/86 Sr)0 values do not correlate with δ 18 O(WR)
values (Fig. 14A), though anomalously low I(Sr)
b0.700 in one sample (Table 5) indicates that the Rb–
Sr isotope system did not remain fully closed. However,
a Rb–Sr isochron for the Katharina complex yields
an age of 593 ± 16 Ma that is consistent with a U–Pb
single zircon age of 583 ± 5 Ma (Katzir et al., 2007),
whereas for the quartz porphyry dykes a Rb–Sr age
could not be calculated.
Fig. 14. Oxygen, Sr and Nd isotope ratios in mafic and felsic rocks
from the bimodal dyke suite: (A) positive correlation between whole
rock δ18O and 87Sr/86Sr(590 Ma); (B) no correlation between whole
rock δ18O and εNd(590 Ma).
In the trachydolerite increase of I(Sr) values is
exhibited less clearly, though the highest measured
value is 0.71241 (Table 5). In mafic rocks Rb is
concentrated in subordinate alkali feldspar while the
main carrier of Sr is plagioclase constituting about 60%
of the rock volume. Since Sr dominates over Rb, alteration of alkali feldspar that may result in resetting of its
87
Rb/86Sr and 87Sr/86Sr ratios would not make a great
impact on the whole–rock 87Sr/86Sr value.
Thus the available radiogenic isotope data showed
that the Rb–Sr isotope characteristics of the dyke rocks
are dominated by late post-crystallization alteration,
hence they cannot be used as petrogenetic indicators.
Low-temperature alteration that caused disturbance
of the isotope systems is still poorly studied in southern
Israel. Its effect on the rock chemical composition was
not yet investigated in detail. However, it was shown
above (Section 7.1) that the composition of felsic dyke
rocks was not markedly changed by alteration processes, though in trachydolerite minor increase of K2O and
Rb contents should be considered.
Y. Katzir et al. / Lithos 97 (2007) 336–364
359
7.5. Origin of rhyolite magmas in the bimodal dyke suite
Radiogenic and stable isotope ratios are commonly
considered as the most reliable tracers of provenance of
magmas. However, in the foregoing sections it was
demonstrated that in the Elat area late post-crystallization processes have extensively altered the magmatic Sr
isotope compositions. Unlike Sr isotope ratios, εNd(T)
values of the bimodal suite vary within fairly narrow
limits, from about + 1.4 to + 3.3 both in felsic and mafic
dyke rocks and show no evidence of correlation with
the whole–rock (87Sr/86Sr)0 and δ18O values (Figs. 8
and 14B). Thus, the alteration processes have not
significantly affected the Sm–Nd isotope system.
The positive εNd(T) values of the mafic dyke
rocks, + 2.0 to +3.3 (Table 5) indicate derivation
from a mildly ‘enriched’ mantle source, similar to those
associated with present day ocean island basalts
(DePaolo, 1988; Stein and Goldstein, 1996). This
source was more enriched relative to the “plume
mantle” that formed at the early stages of the evolution
of the ANS and was transformed into juvenile crust and
lithospheric mantle during the time interval of ∼ 870–
650 Ma (Stein, 2003). According to Stein (2003), the
εNd value of the “plume mantle”, + 5 ± 1, did not
change much during a long period, from 800 Ma to
200 Ma, owing to its low Sm/Nd ratio.
In the felsic dyke rocks the εNd(T) values vary
from + 0.9 to + 3.1. It is likely that in places the Nd
isotope system was slightly affected by alteration. For
example, εNd(T) values of three samples collected from
a single quartz–porphyry dyke (Fig. 1, dyke #8)
are + 2.3, + 2.4, and +0.9 (Table 5, samples A18, A663, and A30). This suggests that the lowest value may be
altered, thus limiting the igneous εNd(T) values of felsic
rocks to the range of +1.4 to +3.1.
Apparently, the complete overlap of Nd isotope ratios
in the coexisting mafic and silicic magmas can be
interpreted in terms of consanguinity of mafic and silicic
melts. However, in the ANS this is not the only possible
interpretation since most of the crust is juvenile and
relatively young (b1 Ga) and thus by the end of the
Neoproterozoic it had not yet developed the radiogenic
isotope signatures characteristic of ancient continental
crust (Stein and Hoffman, 1992; Stern and Kröner,
1993; Furnes et al., 1996; Stein and Goldstein, 1996;
Stern and Abdelsalam, 1998; Stern, 2002; Mushkin
et al., 2003). In particular, it can be seen in the εNd(T) vs
age diagram (Fig. 15) that two sources of alkaline
rhyolite dykes are equally credible in southern Israel.
The first one is coeval mafic magma evolved by
fractional crystallization to form rhyolite. The second
Fig. 15. εNd(T ) vs age for Neoproterozoic igneous rocks and selected
metamorphic rocks of southern Israel. Sources of data: Elat area, this
paper (Table 5), unpublished authors' data for Elat granitic gneiss, and
Stein and Goldstein (1996); Amram area, Mushkin et al. (2003); Timna
area, Beyth et al. (1994).
viable source is older metamorphic basement (granitic
gneiss, crystalline schist, as well as igneous mafic
rocks), which suggests formation of the rhyolite magma
by partial melting of crustal material.
Regardless of the juvenile character of the crust,
oxygen isotopes are a sensitive tracer of upper crustal
contribution to magmas (King et al., 1998; Peck et al.,
2000; Cavosie et al., 2005; Valley et al., 2005). Oxygen
isotope ratios of refractory minerals reveal significant
difference between the mafic and felsic dykes. Whereas
clinopyroxene from the trachydolerite is characterized
by a typical mantle δ 18 O value of 5.8‰, quartz
porphyry from the same composite dyke (dyke #2)
contains zircon with δ18O = 6.45‰, which appreciably
differs from mantle zircon values, 5.3 ± 0.6‰ (2σ;
Valley, 2003). Other quartz porphyries from the bimodal
suite have either similar (6.5‰) or higher (7.2‰) nonmantle δ18O(Zrn) values (Fig. 7; Table 4). It follows that
the coexisting mafic and silicic magmas of the bimodal
suite differ by origin: trachydolerite evolved from
mantle-derived melts, while quartz porphyry had a
significant crustal component in its source.
A-type silicic magma occurs within the bimodal dyke
suite in southern Israel, southwestern Jordan and Sinai,
Egypt. Previous models for the origin of A-type silicic
magma in the northern ANS included fractional
crystallization of mantle-derived magma (Stern and
Voegeli, 1995; Kessel et al., 1998; Jarrar et al., 2003)
and melting of mafic lower crust (Jarrar et al., 1992,
2004). The new oxygen isotope data show that pure
mantle origin of silicic magma is not possible, and a
contribution of upper crustal, high-δ18O sources to the
rhyolite dyke magma is required. Therefore, if melting
360
Y. Katzir et al. / Lithos 97 (2007) 336–364
of lower crust is considered, such crust should have
included a significant component of protoliths that had
been altered near the surface at low temperatures. A
plethora of surface-derived rocks from earlier evolutionary stages of the ANS occur in southern Israel and
should be considered as potential sources for the
rhyolite magma. For example, melting of the peraluminous Elat granite thought to be derived from metasedimentary sources can account for the observed
isotope ratios of the quartz porphyry [δ 18 O(Zrn)
∼ 7‰; Fig. 16; Eyal et al., 2004]. Melting of the late
Proterozoic metamorphic upper crust represented by the
Elat granitic gneiss [δ18O(Zrn) ∼ 7‰; Fig. 16] is also a
possible scenario, which is supported by the Nd isotope
data (Fig. 15). Thus, the role of mantle-derived magma
in generating the quartz porphyry dykes was likely
restricted to advection of heat with only subdued
contribution of mafic material to the magma.
The recently acquired isotope data on the dykes of
the bimodal suite and on some A-type granite plutons in
the northern ANS give a new perspective on the long
standing problem of the origin of the A-type granites
(Loiselle and Wones, 1979; Collins et al., 1982; Whalen
et al., 1987; Eby, 1990; Creaser et al., 1991; Turner
et al., 1992; Kerr and Fryer, 1993; Wickham et al., 1995;
Patiño Douce, 1997; Litvinovsky et al., 1999, 2002; Wu
et al., 2002). In addition to the quartz porphyry dykes,
appreciable crustal contribution is also indicated by the
δ18O (Zrn) values of the alkaline Yehoshafat granite,
6.52 to 6.76‰ (Fig. 16; Table 7). The granite formed in
the beginning of the post-orogenic stage (Table 7) and is
intruded by the bimodal dyke suite. In contrast to the
relatively high oxygen isotope ratios of the rhyolitic
Fig. 16. Oxygen isotope ratios of zircon from igneous felsic rocks and
granitic gneiss in southern Israel and the Sinai Peninsula (the Katharina
ring complex). Two groups of A-type granites with different isotope
signatures are distinguished. Mantle values of δ18O(Zrn)= 5.3 ± 0.6 (2 σ;
Valley, 2003).
dykes and Yehoshafat granite, other A-type granites in
the northern ANS bear a dominant mantle δ 18 O
signature. In particular, the Timna alkaline granite
pluton exposed ∼ 20 km to the north of the study area
and also intruded by the bimodal dyke suite has average
δ18O(Zrn) of 5.52‰ (Table 7). Similar δ18O(Zrn) values
are obtained for syenogranite (5.81‰) and ring dyke
perthite quartz syenite (5.48–5.66‰) from the alkaline
Katharina ring complex (S. Sinai, Egypt) (Table 7).
These values indicate that the generation of the alkaline
granite magma involved either differentiation of mantlederived alkaline mafic magma with minor assimilation
of crustal material (Beyth et al., 1994) or melting of
lower crust formed by underplating of K-rich basalt
magma.
The abundant bimodal dyke swarms, A-type granite
plutons and bimodal volcanic suites formed in an
extentional tectonic environment (e.g., Stern, 1994;
Garfunkel, 2000; Genna et al., 2002). Extensioninduced ascent of mantle-derived mafic magma into
the crust could either melt the existing continental crust
and mix with the partial melts or merely result in
fractional crystallization of mafic magma leading to
granite formation (Eby, 1992; Black and Liégeois,
1993). The presence of two isotopically distinct groups
of post-orogenic alkaline granite in the northern ANS,
suggests that A-type magmatism may originate by
alternative processes almost coevally.
8. Conclusions
1. The Late Pan-African bimodal dyke suite in southern
Israel was formed by injection of alkaline rhyolitic and
trachybasaltic magmas. Formation of composite dykes
points to coexistence of silicic and mafic magmas.
2. Evidence of magma mixing is established. Mixing
occurred both at shallow depth, in the process of
formation of composite dykes, and at greater depth,
still prior to dyke emplacement.
3. Dyke rocks have been subjected to extensive lowtemperature alteration that caused significant disturbance of the magmatic 18O/16O ratios in alkali
feldspar and plagioclase. Magmatic 18O/16O ratios in
zircon, clinopyroxene and partly in quartz have been
retained.
4. The Rb–Sr isotope ratios of the quartz porphyry are
also dominated by low-temperature hydrous alteration process; hence they do not characterize the
magmatic process.
5. Initial magmatic Nd isotope ratios were preserved in
most rocks, and the εNd(T) values in mafic and felsic
rocks overlap and range from ∼ 1.2 to 2.9. However,
Y. Katzir et al. / Lithos 97 (2007) 336–364
two sources of rhyolite melt are equally credible: Krich basalt magma that produced coeval mafic dykes,
and metamorphic rocks from the juvenile crust.
Oxygen isotope ratios of zircon in quartz porphyry
reveal significant crustal contribution to the rhyolites,
suggesting that mafic and A-type silicic magmas are
coeval, but not co-genetic.
6. Two groups of Late Neoproterozoic A-type granites
are distinguished in the region: (1) those produced
mostly from a mantle-derived source, with δ18O(Zrn)
ranging from 5.5 to 5.7‰ (Timna and Katharina
granitoids), and (2) those with significant contribution
of the modified juvenile crustal component, with δ18O
(Zrn) varying from 6.5 to 7.2‰ (Elat quartz porphyry
dykes and Yehoshafat granite). This attests that A-type
silicic magmas may originate by alternative processes
almost coevally.
Acknowledgements
This study was supported by the Ministry of National
Infrastructure and Energy, Israel (grant #23-17-04), the
Israel Science Foundation (grant 142/02) and by the
National Science Council (NSC), Taiwan (grants
NSC93-2116-M-001-025, NSC94-2116-M-001-021,
NSC94-2752-M-002-010-PAE). Authors are greatly
indebted to M. Beyth and an anonymous reviewer for
valuable recommendations and corrections that provided a significant improvement to the paper.
Appendix A. Supplementary data
Supplementary data associated with this article
can be found, in the online version, at doi:10.1016/j.
lithos.2007.01.004.
References
Agron, N., Bentor, Y.K., 1981. The volcanic massif of Biq'at Hayareah
(Sinai–Negev), a case of potassium metasomatism. J. Geol. 89,
479–496.
Baer, G., Beyth, M., Reches, Z., 1994. Dyke emplaced into fractured
basement, Timna Igneous Complex, Israel. J. Geophys. Res. 99,
24039–24050.
Baker, D.R., 1989. Tracer versus trace element diffusion: diffusional
decoupling of Sr concentration fron Sr isotope composition.
Geochim. Cosmochim. Acta 53, 3015–3023.
Baker, D.R., 1990. Chemical interdiffusion of dacite and rhyolite:
anhydrous measurements at 1 atm and 10 kbar, application of
transition state theory, and diffusion in zoned magma chambers.
Contrib. Mineral. Petrol. 104, 407–423.
Bentor, Y.K., 1985. The crust evolution of the Arabo-Nubian Massif
with special reference to the Sinai Peninsula. Precambrian Res.
28, 1–74.
361
Bentor, Y.K., Eyal, M., 1987. The geology of southern Sinai, its
implication for the evolution of the Arabo-Nubian Massif. Isr.
Acad. Sci. Humanit., Jerusalem.
Beyth, M., Stern, R.J., Altherr, R., Kröner, A., 1994. The late
Precambrian Timna igneous complex, Southern Israel: Evidence
for comagmatic-type sanukitoid monzodiorite and alkali granite
magma. Lithos 31, 103–124.
Beyth, M., Longstaffe, F.J., Ayalon, A., Matthews, A., 1997.
Epigenetic alteration of the Precambrian igneous complex at
Mount Timna, southern Israel: oxygen-isotope studies. Isr. J. Earth
Sci. 46, 1–11.
Beyth, M., Heimann, A., 1999. The youngest igneous event in the
crystalline basement of the Arabian–Nubian Shield, Timna
Igneous Complex. Isr. J. Earth Sci. 48, 113–120.
Bielski, M., 1982. Stages in the evolution of the Arabo-Nubian massif
in Sinai. Ph.D. thesis, Hebrew Univ., Jerusalem (in Hebrew;
English Abstract, Tables, Figures).
Black, R., Liégeois, J.P., 1993. Cratons, mobile belts, alkaline rocks
and continental lithospheric mantle: the Pan-African testimony.
J. Geol. Soc. (Lond.) 150, 89–98.
Carrigan, C.R., Eichelberg, J.C., 1990. Zoning of magmas by viscosity
in volcanic conduits. Nature 343, 248–251.
Cavosie, A.J., Valley, J.W., Wilde, S.A., E.I.M.F., 2005. Magmatic 18O
in 4400–3900 Ma detrital zircons: a record of the alteration and
recycling of crust in the early Archean. Earth Planet. Sci. Lett. 235,
663–681.
Chiba, H., Chacko, T., Clayton, R.N., Goldsmith, J.R., 1989. Oxygen
isotope fractionations involving diopside, forsterite, magnetite and
calcite: application to geothermometry. Geochim. Cosmochim.
Acta 53, 2895–2995.
Clayton, R.N., Goldsmith, J.R., Mayeda, T.K., 1989. Oxygen isotope
fractionation in quartz, albite, anorthite and calcite. Geochim.
Cosmochim. Acta 53, 725–733.
Collins, W.J., Beams, S.D., White, A.J.R., Chappell, B.W., 1982.
Nature and origin of A-type granites with particular reference to
Southeastern Australia. Contrib. Mineral. Petrol. 80, 189–200.
Creaser, R.A., Price, R.C., Wormald, R.J., 1991. A-type granites revisited:
assessment of a residual-source model. Geology 19, 63–166.
Criss, R.E., Taylor, H.P., 1983. An 18O/16O and D/H study of Tertiary
hydrothermal systems in the southern half of the Idaho batholith.
Geol. Soc. Amer. Bull. 94, 640–663.
Criss, R.E., Taylor, H.P., 1986. Meteoric-hydrothermal systems. In: Valley,
J.W., Taylor, H.P., O'Neil, J.R. (Eds.), Stable Isotopes in High
Temperature Geological Processes. Reviews in Mineralogy, vol. 16.
Mineral. Soc. America, Washington, DC, pp. 373–424.
DePaolo, D.J., 1988. Neodynium Isotope Geochemistry: An Introduction. Springer-Verlag, New York.
Eby, G.N., 1990. The A-type granitoids: a review of their occurrence
and chemical characteristics and speculations on their petrogenesis. Lithos 26, 115–134.
Eby, G.N., 1992. Chemical subdivision of the a-type granitoids:
petrogenetic and tectonic implications. Geology 20, 641–644.
Eiler, J.M., Farley, K.A., Valley, J.W., Hofmann, A.W., Stolper, E.M.,
1996. Oxygen isotope constraints on the sources of Hawaiian
volcanism. Earth Planet. Sci. Lett. 144, 453–468.
Eiler, J.M., Crawford, A., Elliott, T., Farley, K.A., Valley, J.W.,
Stolper, E.M., 2000. Oxygen isotope geochemistry of oceanic-arc
lavas. J. Petrol. 41, 229–256.
Eyal, M., Hezkiyahu, T., 1980. Katharina Pluton — the outlines of a
petrological framework. Isr. J. Earth Sci. 29, 41–52.
Eyal, Y., Eyal, M., Kröner, A., 1991. Geochronology of the Elat
Terrain metamorphic basement, and its implication for crustal
362
Y. Katzir et al. / Lithos 97 (2007) 336–364
evolution of the NE part of the Arabian–Nubian Shield. Isr.
J. Earth Sci. 40, 5–16.
Eyal, M., Litvinovsky, B.A., Katzir, Y., Zanvilevich, A.N., 2004. The
Pan-African high-K calc-alkaline peraluminous Elat granite of
southern Israel: geology, geochemistry and petrogenesis. J. Afr.
Earth Sci. 40, 115–136.
Furman, T., Spera, F.J., 1985. Co-mingling of acid and basic magma
with the implications for the origin of mafic I-type xenoliths: field
and petrochemical relations of an unusual dike complex at Eagle
Lake, Sequoia National Park, California, USA. J. Volcanol.
Geotherm. Res. 24, 151–178.
Furnes, H., El-Sayed, M.M., Khalil, S.O., Hassanen, M.A., 1996. PanAfrican magmatism in the Wadi El-Imra district, Central Eastern
Desert, Egypt: geochemistry and tectonic environment. J. Geol.
Soc. (Lond.) 153, 705–718.
Gal, A., 2006. Oxygen Isotope Perspective on Post-Orogenic Alkaline
Magmatism in the Arabian–Nubian Shield of Southern Israel. MS
thesis, Ben–Gurion University of the Negev, Be'er Sheva, Israel,
(in Hebrew; English Abstract).
Garfunkel, Z., 2000. History and paleogeography during the PanAfrican orogen to stable platform transition: Reappraisal of the
evidence from Elat area and the northern Arabian–Nubian Shield.
Isr. J. Earth Sci. 48, 135–157.
Garfunkel, Z., Eyal, Y., Eyal, M., Weissbrod, T., Bakler, N.,
Shimron, A.E., Peltz, S., Gutkin, V., Bartov, Y., Drukman, Y.,
Rozenfeld, M., Sneh, A., 2000. Geological map of the Northern
Gulf of Elat area, scale 1:100,000 (sheet 26 Elat). Geol. Surv.
Isr., Jerusalem.
Genna, A., Nehlig, P., Le Goff, E., Guerrot, C., Shanti, M., 2002.
Proterozoic tectonism of the Arabian Shield. Precambrian Res.
117, 21–40.
Gibson, I.L., Walker, G.P.L., 1963. Some composite rhyolite/basalt
lavas and related composite dikes in eastern Iceland. Proc. Geol.
Assoc. 74, 301.
Gutkin, V., Eyal, Y., 1998. Geology and evolution of Precambrian
rocks, Mt. Shelomo, Elat area. Isr. J. Earth Sci. 47, 1–18.
Iacumin, M., Mazaroli, A., El-Metwally, A.A., Piccirillo, E.M.,
1998. Neoproterozoic dyke swarms from southern Sinai (Egypt):
geochemistry and petrogenetic aspects. J. Afr. Earth Sci. 26,
49–64.
Jarrar, G., 2001. The youngest Neoproterozoic mafic dike suite in the
Arabian Sield: mildly alkaline dolerites from South Jordan — their
geochemistry and petrogenesis. Geol. Mag. 138, 309–323.
Jarrar, G., Wachendorf, H., Saffarini, G., 1992. A Late Proterozoic
bimodal volcanic/subvolcanic suite from Wadi Araba, southwest
Jordan. Precambrian Res. 56, 51–72.
Jarrar, G., Stern, R.J., Saffarini, G., Al-Zubi, H., 2003. Late and postorogenic Neoproterozoic intrusions of Jordan: implications for
crustal growth in the northernmost segment of the East African
Orogen. Precambrian Res. 123, 295–319.
Jarrar, G., Saffarini, G., Baumann, A., Wachendorf, H., 2004. Origin,
age and petrogenesis of Neoproterozoic composite dikes from the
Arabian–Nubian Sield, SW Jordan. Geol. J. 39, 157–178.
Kanaris-Sotiriou, R., Gill, F.G.F., 1985. Hybridization and the
petrogenesis of composite intrusions: The dyke at An Cumhann,
Isle of Arran, Scotland. Geol. Mag. 122, 361–372.
Katzir, Y., Litvinovsky, B.A., Eyal, M., Zanvilevich, A.N., Vapnik, Ye.,
2006. Four successive episodes of Late Pan-African dykes in the
central Elat area, southern Israel. Isr. J. Earth Sci. 55 (2).
Katzir, Y., Eyal, M., Litvinovsky, B.A., Jahn, B-M., Zanvilevich, A.N.,
Valley, J.W., Beeri, Y., Pelly, I., Shimshilashvili, E., 2007.
Petrogenesis of A-type granites and origin of vertical zoning in
the Katharina pluton, area of Gebel Mussa (Mt. Moses), Sinai,
Egypt. Lithos. 95, 208–228.
Kerr, A., Fryer, B.J., 1993. Nd isotope evidence for crust–mantle
interaction in the generation of A-type granitoid suites in Labrador,
Canada. Chem. Geol. 104, 39–60.
Kessel, R., Stein, M., Navon, O., 1998. Petrogenesis of late
Neoproterozoic dikes in the northern Arabian–Nubian Shield.
Implication for the origin of A-type granites. Precambrian Res. 92,
195–213.
Keto, L.S., Jacobsen, S.B., 1987. Nd and Sr isotopic variations of
Early Paleozoic oceans. Earth Planet. Sci. Lett. 84, 27–41.
King, E.M., Valley, J.W., Davis, D.W., Edwards, G.R., 1998. Oxygen
isotope ratios of Archean plutonic zircons from granite–greenstone
belts of the Superior Province: indicator of magmatic source.
Precambrian Res. 92, 365–387.
Koyaguchi, T., Takada, A., 1994. An experimental study of the
formation of composite intrusions from zoned magma chambers.
J. Volcanol. Geotherm. Res. 59, 261–267.
Lasher, C.E., 1990. Decoupling of chemical and isotopic exchange
during magma mixing. Nature 344, 235–237.
Lasher, C.E., 1994. Kinetics of Sr and Nd exchange in silicate liquids:
theory, experiments, and applications to uphill diffusion, isotopic
equilibration and irreversible mixing of magmas. J. Geophys. Res.
99, 9585–9604.
Liégeois, J.-P., Black, R., 1987. Alkaline magmatism subsequent to
collision in the Pan-African belt of the Adrar des Iforas. In: Fitton,
J.G., Upton, B.G.J. (Eds.), Alkaline Igneous Rocks. Geol. Soc.
Spec. Publ., vol. 30, pp. 381–401.
Litvinovsky, B.A., Zanvilevich, A.N., Kalmanovich, M.A., 1995a.
Recurrent mixing and mingling of coexisting syenite and basalt
magmas in the Ust'-Khilok massif, Transbaikalia, and its
petrologic significance. Petrology 3, 115–137.
Litvinovsky, B.A., Zanvilevich, A.N., Lyapunov, S.M., Bindeman, I.N.,
Davis, A.M., Kalmanovich, M.A., 1995b. Model of composite
basite-granitoid dike generation (Shaluta Pluton, Transbaikalia).
Russ. Geol. Geophys. 36, 1–19.
Litvinovsky, B.A., Zanvilevich, A.N., Wickham, S.M., Steele, I.M.,
1999. Origin of syenitic magmas from anorogenic granitoid series:
syenite–granite series from Transbaikalia. Petrology (Moscow) 7,
483–508.
Litvinovsky, B.A., Jahn, B.-M., Zanvilevich, A.N., Saunders, A.,
Poulain, S., 2002. Petrogenesis of syenite–granite suites from
the Bryansky Complex (Transbaikalia, Russia): implications for
the origin of A-type granitoid magmas. Chem. Geol. 189,
105–133.
Litvinovsky, B.A., Eyal, M., Zanvilevich, A.N., Katzir, Y., Jahn B.-M.,
Beeri, Y., Valley, J.W., Vapnik, Ye., in press. Late Neoproterozoic
dykes in the northernmost Arabian–Nubian shield (southern
Israel): sequence of dyke episodes, geochemical and isotope
evidence of magma sources. Submitted to Int. J. Earth Sci.
Loiselle, M.C., Wones, D.R., 1979. Characteristics and origin of
anorogenic granites. Geol. Soc. Am. Abstr. Progr. 11, 468.
Macpherson, C.G., Gamble, J.A., Mattey, D.P., 1998. Oxygen isotope
geochemistry of lavas from an oceanic to continental arc transition,
Kermadec–Hikurangi margin, SW Pacific. Earth Planet. Sci. Lett.
160, 609–621.
Marco, S., Ron, H., Matthews, A., Beyth, M., Navon, O., 1993.
Chemical remanent magnetism related to the Dead Sea Rift:
Evidence from Precambrian igneous rock of mount Timna,
southern Israel. J. Geophys. Res. 98, 16001–16012.
Marian, P.D., Piccoli, P.P.M., 1989. Tectonic discrimination of
granitoids. Geol. Soc. Amer. Bull. 101, 635–643.
Y. Katzir et al. / Lithos 97 (2007) 336–364
Marshall, L.A., Sparks, R.S.J., 1984. Origin of some mixed-magma
and net-veined ring intrusions. J. Geol. Soc. (Lond.) 141, 171–182.
Mattey, D., Lowry, D., Macpherson, C., 1994a. Oxygen isotope
composition of mantle peridotite. Earth Planet. Sci. Lett. 128, 231–241.
Mattey, D., Lowry, D., Macpherson, C., Chazot, G., 1994b. Oxygen
isotope composition of mantle minerals by laser fluorination
analysis: homogeneity in peridotites, heterogeneity in eclogites.
Mineral. Mag. 58A, 573–574.
Matthews, A., Ayalon, A., Ziv, A., Shaked, J., 1999. Hydrogen and
oxygen isotope studies of alteration in the Timna igneous complex.
Isr. J. Earth Sci. 48, 121–131.
Moghazi, A.M., 2002. Petrology and geochemistry of Pan-African
granitoids, Kab Amiri area, Egypt — imolications for tectonomagmatic
stages of the Nubian Shield evolution. Mineral. Petrol. 75, 41–67.
Mushkin, A., Navon, O., Halicz, L., Heimann, A., Wöerner, G., Stein,
M., 1999. Geology and geochronology of the Amram Massif
rocks, southern Negev Desert, Israel. Isr. J. Earth Sci. 48, 179–193.
Mushkin, A., Navon, O., Halicz, L., Heimann, A., Hartmann, G.,
Stein, M., 2003. The petrogenesis of A-type magmas from the
Amram Massif, southern Israel. J. Petrol. 44, 815–832.
Page, F.Z., Deangelis, M., Fu, B., Kita, N., Lancaster, P., Valley J.W.,
2006. Slow oxygen diffusion in zircon.16th V.M. Goldschmidt
Conference, Melbourne, Australia. Absracts.
Patiño Douce, A.E., 1997. Generation of metaluminous A-type
granites by low-pressure melting of calc-alkaline granitoids.
Geology 25, 743–746.
Pearce, J.A., Cann, J.R., 1973. Tectonic setting of basic volcanic rocks
determined using trace element analyses. Earth Planet. Sci. Lett.
19, 290–300.
Pearce, J.A., Harris, N.B., Tindle, A.G., 1984. Trace element
discrimination diagrams for the tectonic interpretation of granitic
rocks. J. Petrol. 25, 956–983.
Peck, W.H., King, E.M., Valley, J.W., 2000. Oxygen isotope
perspective on Precambrian crustal growth and maturation.
Geology 28, 363–366.
Poli, G.E., Tommazini, S., 1991. Model of the origin and significance
of microgranular enclaves in calc-alkaline granitoids. J. Petrol. 32,
657–666.
Snyder, D., Crambles, C., Tait, S., Wiebe, R.A., 1997. Magma
mingling in dikes and sills. J. Geol. 105, 75–86.
Snyder, D., Tait, S., 1998. The imprint of basalt on the geochemistry of
silicic magmas. Earth Planet. Sci. Lett. 160, 433–445.
Spicuzza, M.J., Valley, J.W., Kohn, M.J., Girard, J.P., Fouillac, A.M.,
1998a. The rapid heating, defocused beam technique: a CO2-laserbased method for highly precise and accurate determination of
δ18O values of quartz. Chem. Geol. 144, 195–203.
Spicuzza, M.J., Valley, J.W., McConnell, V.S., 1998b. Oxygen isotope
analysis of whole rock via laser fluorination: an air-lock approach.
Geol. Soc. Am. Abstr. Progr. 30, 80.
Stein, M., 2003. Tracing the plume material in the Arabian–Nubian
Shield. Precambrian Res. 123, 223–234.
Stein, M., Hoffman, A.W., 1992. Fossil plumes beneath the Arabian
lithosphere? Earth Planet. Sci. Lett. 114, 193–209.
Stein, M., Goldstein, S.L., 1996. From plume head to continental
lithosphere in the Arabian–Nubian shield. Nature 382, 773–778.
Stein, M., Navon, O., Kessel, R., 1997. Chromatographic metasomatism of the Arabian–Nubian lithosphere. Earth Planet. Sci. Lett.
152, 75–91.
Stern, R.J., 1994. Arc assembly and continental collision in the
Neoproterozoic East African orogen. Implications for the consolidation of Gondwanaland. Annu. Rev. Earth Planet. Sci. 22,
319–351.
363
Stern, R.J., 2002. Crustal evolution in the East African Orogen: a
neodymium isotopic perspective. J. Afr. Earth Sci. 34, 109–117.
Stern, R.J., Gottfried, D., Hedge, C.E., 1984. Late Precambrian rifting
and crustal evolution in the North Eastern Desert of Egypt.
Geology 12, 168–172.
Stern, R.J., Hedge, C.E., 1985. Geochronologic and isotopic
constraints on Late Precambrian crustal evolution in the Eastern
Desert of Egypt. Am. J. Sci. 285, 97–127.
Stern, R.J., Gottfried, D., 1986. Petrogenesis of the Late Precambrian
(575–600 Ma) bimodal suite in the Northeast Africa. Contrib.
Mineral. Petrol. 92, 492–501.
Stern, R.J., Voegeli, D.F., 1987. Geochemistry, geochronology, and
petrogenesis of a Late Precambrian (= 590 Ma) composite dike
from the North Eastern Desert of Egypt. Geol. Rundsch. 76,
325–341.
Stern, R.J., Sellers, G., Gottfried, D., 1988. Bimodal dike swarms in
the North of the Eastern Desert of Egypt: Significance for the
origin of the Precambrian “A-type” Granites in Northern AfroArabia. In: Greiling, R., Gaby, R. (Eds.), The Pan-African Belt of
NE Africa and Adjacent Areas: Tectonic Evolution and
Economic Aspects. Frieder. Vieweg. and Sohn, Braunschweig,
Germany, pp. 147–179.
Stern, R.J., Kröner, A., 1993. Late Precambrian crustal evolution in
NE Sudan, isotopic and geochronologic constraints. J. Geol. 101,
555–574.
Stern, R.J., Voegeli, D., 1995. Late Neoproterozoic (570 Ma)
composite dykes from the Northern Arabian–Nubian Shield:
Petrogenetic perspectives on contemporaneous mafic and felsic
igneous activity. Third international Dyke Conference, Jerusalem,
September 4–8, 1995, p. 69. Abstracts.
Stern, R.J., Abdelsalam, M.G., 1998. Formation of continental crust in
the Arabian–Nubian Shield: evidence from granitic rocks of the
Nakasib suture, NE Sudan. Geol. Rundsch. 87, 150–160.
Stoeser, D.B., Camp, V.E., 1985. Pan-African microplate accretion of
Arabian shield. Geol. Soc. Amer. Bull. 96, 817–826.
Sun, S., McDonough, W.F., 1989. Chemical and isotopic systematics
of oceanic basalts: implications for mantle compositions and
processes. In: Saunders, A.D., Norry, M.J. (Eds.), Magmatism in
the Ocean Basins. Geol. Soc. Spec. Publ., vol. 42, pp. 313–345.
Sylvester, P.J., 1989. Post-collisional alkaline granites. J. Geol. 97,
261–281.
Titov, A.V., Litvinovsky, B.A., Zanvilevich, A.N., Shadaev, M.G.,
2000. Hybridization in composite basic-rock-leucogranite dikes of
the Ust'-Khilol massif, Transbaikalia. Russ. Geol. Geophys. 41,
1662–1675.
Turner, S.P., Foden, J.D., Morrison, R.S., 1992. Derivation of some Atype magmas by fractionation of basaltic magma: an example from
the Padthway Ridge, South Australia. Lithos 28, 151–179.
Valley, J.W., 2003. Oxygen isotopes in zircon. In: Hanchar, J.M.,
Hoskin, W.O. (Eds.), Zircon. Reviews in Mineralogy and
Geochemistry, vol. 53. Mineral. Soc. America, Washington, DC,
pp. 343–385.
Valley, J.W., Chiarenzelli, J.R., McLelland, J.M., 1994. Oxygen
isoitope geochemistry of zircon. Earth Planet. Sci. Lett. 126,
187–206.
Valley, J.W., Kitchen, N., Kohn, M.J., Niendorf, C.R., Spicuzza, M.J.,
1995. UWG-2, a garnet standard for oxygen isotope ratios:
strategies for high precision and accuracy with laser heating.
Geochim. Cosmochim. Acta 59, 5523–5531.
Valley, J.W., Bindeman, I.N., Peck, W.H., 2003. Empirical calibration
of oxygen isotope fractionation in zircon. Geochim. Cosmochim.
Acta 67, 3257–3266.
364
Y. Katzir et al. / Lithos 97 (2007) 336–364
Valley, J.W., Lackey, J.S, Cavosie, A.J., Clechenko, C.C., Spicuzza,
M.J., Basei, M.A.S., Bindeman, I.N., Ferreira, V.P., Sial, A.N.,
King, E.M., Peck, W.H., Sinha, A.K., Wei, C.S., 2005. 4.4 billion
years of crustal maturation: oxygen isotopes in magmatic zircon.
Contrib. Mineral. Petrol. 150, 561–580.
Vogel, T.A., Wilband, J.T., 1978. Coexisting acidic and basic melts:
geochemistry of a composite dike. J. Geol. 86, 353–371.
Wachendorf, H., Zachmann, D., Jarrar, G.H., 1985. The role of
pressure in control of potassium, sodium, and copper concentration
in hypabyssal intrusives as demonstrated in Late Precambrian
dikes in southern Jordan. Precambrian Res. 30, 221–248.
Wenner, D.B., Taylor Jr., 1976. Oxygen and hydrogen isotope studies
of Precambrian granite–rhyolite terrane, St. Francois Mountains,
southeastern Missouri. Geol. Soc. Amer. Bull. 87, 1587–1598.
Whalen, J.B., Currie, K.L., Chappe, B.W., 1987. A-type granites:
geochemical characteristics, discrimination and petrogenesis.
Contrib. Mineral. Petrol. 95, 407–419.
Wickham, S.M., Litvinovsky, B.A., Zanvilevich, A.N., Bindeman, I.N.,
1995. Geochemical evolution of Phanerozoic magmatism in
Transbaikalia, East Asia: a key constraint of the origin of K-rich
silicic magmas and the process of cratonization. J. Geophys. Res.
100/B8, 15641–15654.
Wiebe, R.A., Ulrich, R., 1997. Origin of composite dikes in the
Gouldsboro granite, coastal Maine. Lithos 40, 157–178.
Winchester, J.A., Floyd, P.A., 1977. Geochemical discrimination of
different magma series and their differentiation products using
immobile elements. Chem. Geol. 70, 319–339.
Wu, F.-Y., Sun, D.-Y., Li, H.-M., Jahn, B.-M., Wilde, S.A., 2002. Atype granites in Northeastern China: age and geochemical
constraints on their petrogenesis. Chem. Geol. 187, 143–173.
Zorpi, M.J., Couloh, C., Orsini, J.B., 1991. Hybridization between
felsic and mafic magmas in calc-alkaline granitoids — a case study
in northern Sardinia, Italy. Chem. Geol. 92 (1/3), 45–86.